Professional Documents
Culture Documents
A Thesis
by
Mohammad Al Amin Siddique
2005
~
11
CONCRETE STRUCTURES
M.Sc. Thesis
by
Mohammad Al Amin Siddique
Roll No.: 040204305 F
A Master of Science (M. Sc.) thesis approved as to the style and content in partial
fulfillment of the requirement for the degree of Master of Science in Civil
Engineeri (Struc ural) on 13th June, 2005.
U.
Dr. Sk. Sekfnder Ali Member (Ex-Officio)
Professor and Head
Department of Civil Engineering
BUET, Dhaka-lOOO.
OJI ~
~b-. ------
Dr. Ahsanu Ka Ir Member
Professor
Department of Civil Engineering
BUET, Dhaka-IOOO.
•
III
DEDICATIONS
To My Family Members
v
ACKNOWLEDGEMENT
Thanks to Almighty Allah for His graciousness, unlimited kindness and with the
blessings of Whom the good deeds are fulfilled.
The author wishes his deepest gratitude to his supervisor Dr. Md. Abdur Rouf,
Professor, Department of Civil Engineering, BUET, Dhaka-1000, for his continuous
guidance, invaluable suggestions, affectionate encouragement, generous help and
unfailing enthusiasm at every stage of this study. His active interest in this topic and
invaluable advice were sources of the author's inspiration.
The author is grateful to Dr. A.F .M. Saiful Amin, Assistant Professor, Department of
Civil Engineering, BUET, for his help in using Visual FORTRAN 6.0 package. The
author is also grateful to his colleagues for their active co-operation.
A very special debt of deep gratitude is offered to his grandfather, parents, sisters,
uncle and aunt who are always a constant source of inspiration throughout his life.
•
IV
DECLARATION
It is hereby declared that except for the contents where specific reference has been
made to the work of others, the studies contained in this thesis are the result of
investigation carried out by the author. No part of this thesis has been submitted to
any other University or other educational establishment for a Degree, Diploma or
other qualification (except for publication).
ABSTRACT
Reinforced concrete (RC) structures are commonly used and are designed to satisfy
criteria of serviceability and safety. To ensure the serviceability it is necessary to
predict the cracking and the deflections of RC members under loads and accurate
estimation of the collapse load is essential to provide margin of safety. The behavior
of RC members is complex due to its constituent materials which led engineers in the
past to rely heavily on empirical formulas, for the design of concrete structures,
derived from numerous experiments: The advent of digital computers and powerful
methods of analysis obviate the need for experiments recognizing that tests are time
consuming and costly. Hence, the development of a nonlinear numerical model is
necessary for the analysis ofRC beams, columns, and frames up to collapse load. The
objective of this study was to develop reliable and computationally efficient finite
element model for the analysis of RC members. The stress-strain behavior of the
reinforcing steel is assumed bilinear which can model elastic perfectly plastic and
elastic with strain hardening. The stress-strain behavior of concrete is considered to be
parabolic. The constituents of a RC section such as concrete and reinforcing steel are
represented by separate material models, which are combined together with the model
of the interaction between concrete and reinforcing steel considering perfect bond to
describe the behavior ofthe section.
A number of correlation studies have been conducted to evaluate the ability of the
numerical model in predicting the response of load-deflection and moment-curvature
behavior of RC members. Comparisons are made on available experimental data and
also on the analytical results. These comparisons consist of 6 experimental results for
RC beams, 2 for RC columns and 3 for RC frames. Analytical investigation has been
made on the RC beams and columns. From analyses using the model it has been
found that the present model captures the load-deflection and moment-curvature
behavior quite well. The material nonlinearity of the RC section has been effectively
used to present the complex behavior of RC members. Effect of tensile strength of
concrete on the ultimate load carrying capacity is insignificant but the strain
hardening of reinforcing steel plays a vital role for the collapse mechanism. The
ductility of a singly reinforced RC section decreases as the tension reinforcement
increases. The presence of compression steel increases the ductility of a RC section
vii
significantly. The flexural response ofHSC beams has also been analyzed considering
the material stress-strain relationship used in the model. Increase of axial load in a RC
column increases the moment carrying capacity up to a certain limit and then the
effect reverses and decreases the ductility of the section. Increase in steel content
increases the load carrying capacity of RC columns but decreases the ductility for the
same concrete strength. The ductility of RC columns increases as the concrete
compressive strength is increased.
\c.,
VIII
Table of Contents
Acknowledgement V
Abstract VI
List of Tables X
List of Figures XI
Chapter 1 INTRODUCTION I
1.1 General I
1.2 Objectives ofthe Study 3
I.3 Scope and Methodology of the Study 3
1.4 Organization of the Study 4
Chapter 2 LITERATURE REVIEW 6
2.1 General 6
2.2 Nonlinear Analysis ofRC Elements 6
2.3 Concluding Remarks II
Chapter 3 DEVELOPMENT OF A NONLINEAR NUMERICAL I3
MODEL -NLRCF
3.1 Introduction I3
3.2 Properties of Constituent Materials 14
3.2.1 General 14
3.2.2 Concrete 16
3.2.2.1 Behavior of Concrete 16
3.2.2.2 Concrete Models 17
3.2.2.3 Behavior of a Cracked RC Section 20
3.2.3 Reinforcing Steel 22
3.2.4 Bond between Reinforcing Steel and Concrete 25
3.3 Finite Element Representation 26
3.4 Stiffness of a Beam Element 29
IX
List of Tables
Table 5.6 Predicted and test loads and types of failure (Ernst et aI., 1973) 81
Xl
List of Figures
List of Notations
a = shear span
b=width
d = depth of the tension steel from top ofthe section
d' = depth of compression steel from top of the section
f = compressive stress in concrete
t; = tensile strength of concrete (modulus of rupture)
f e = crushing strength of concrete
fy = yield strength of steel reinforcement
fu = ultimate strength of steel reinforcement
h= depth of the section
As = area of tension reinforcement
A's = area of compression reinforcement
E1 = flexural rigidity
Ee = modulus of elasticity of concrete
Esl = Es = elastic modulus of steel reinforcement
Es2 = Esh = plastic modulus of steel reinforcement
{F} = load vector
[K] = stiffuess matrix
L = total depth of the section/2
P = axial load
Po = axial load capacity of the section without flexure
M=moment
Mu = ultimate bending moment
Em = strain in concrete corresponds to crushing strength of concrete
Ey = strain in steel corresponds to yield strength of steel reinforcement
Eu = strain in steel corresponds to ultimate strength of steel reinforcement
ED = EF = convergence tolerance
<p = curvature
p = percentage of tension reinforcement of the section (As/bd)
p' = percentage of compression reinforcement of the section (A's/bd)
Pb = reinforcement ratio producing balanced strain condition in a section
,
XIV
List of Abbreviations
FE = Finite Element
HSC =High Strength Concrete
NSC = Normal Strength Concrete
PC = Prestressed Concrete
RC = Reinforced Concrete
NLRCF = Nonlinear numerical model for analyzing Reinforced Concrete Frame
I
Chapter 1
INTRODUCTION
1.1 General
Reinforced concrete (Re) has become one of the most important building materials
and is widely used in many types of engineering structures. The economy, the
efficiency, the strength and the stiffuess of reinforced concrete make it an attractive
material for a wide range of structural applications. The ultimate objective of the
designer is to create a structure that is safe and economical. The safety and
serviceability assessment of the structures necessitate the development of accurate and
reliable methods and models for their analysis. The rise in cost of materials used in
structures and labor costs encourage engineers to seek more economical alternative
designs often resorting to innovative construction methods but without lowering the
safety of the structure. In addition, the extent and impact of disaster in terms of human
and economical loss in the event of structural failure promote designers to check the
design thoroughly. As a result, careful and detailed structural safety analysis becomes
more and more necessary. The objective of such an analysis is the investigation ofthe
behavior of the structure under all possible loading conditions, both, monotonic and
cyclic, its time-dependent behavior, and, especially, its behavior under overloading
and upto load near collapse.
For the development of advanced design and analysis of structures the need for
experimental research continues. Experiments provide a firm basis for the design a
2
These complex phenomena have led engineers in the past to rely heavily on empirical
formulas for the design of concrete structures, which were derived from numerous
experiments. With the advent of digital computers and powerful methods of analysis,
such as the finite element method, much effort to develop analytical solutions which
would obviate the need for experiments have been undertaken by investigators. The
finite element method has thus become a powerful computational tool which allows
complex analyses of the nonlinear response of RC structures to be carried out in a
routine fashion and is widely used. With this method the importance and interaction
of different nonlinear effects on the response of RC structures can be studied
analytically. Commercial softwares, however, generally not yet able to
comprehensively take into consideration all such effects. Hence, the development of a
3
nonlinear numerical model is necessary for the analysis of reinforced concrete beams,
columns, and frames up to col1apse load.
!J
\'t
4
numerical model for masonry has been developed by Rouf (1985). To develop the
numerical model for RC structures, the stiffness properties described above have been
adopted here too.
There are four basic steps involved in nonlinear analysis. These are the formation of
the current stiffness matrix, the solution of the equilibrium equations for the
displacement increments, the determination of force equilibrium of all elements in the
model and the convergence check. Since the global stiffness matrix of the structure
depends on the displacement increments, the solution of equilibrium equations is
typically accomplished with an iterative method through convergence check. The
nonlinear problem has been solved by the modified Newton-Raphson approach. The
stiffness matrix has been updated once in each load increment and it was left constant
for rest ofthe iteration.
This study is organized according to the stages followed for the development of the
nonlinear numerical model. Thus, Chapter I introduces a general statement of the
problem and the objectives of the research. Chapter 2 reviews the available literature
discussing various studies conducted of nonlinear numerical modeling on reinforced
concrete members. Chapter 3 describes the development of nonlinear numerical
o
o
5
model for RC beam, column and frame. There is flexibility in the model to adopt
either material nonlinearity or combination of material and geometric nonlinearities.
However, for RC frames only material nonlinearity has been considered. This chapter
also focuses the behavior of steel and concrete in RC structural elements and their
interaction in a RC member, development of stiffuess matrix and nonlinear solution
technique such as modified Newton-Raphson method which is used for solving the
nonlinear problems in RC members and convergence criteria. Chapter 4 presents the
results obtained from analyses using the proposed model showing the load-deflection,
moment-curvature relationship, effect of steel percentages on the behavior of beams,
failure pattern due to variation of steel ratio, effect of tensile strength of concrete,
strain hardening behavior of steel. Chapter 5 presents the results obtained from
analyses using the present model showing the ultimate strength and deflection
predictions of RC columns and load-deflection responses of RC frames. Verification
of the results of the present numerical model has been done by comparing the
numerical results with the existing experimental data. Theoretical investigation has
been carried out to investigate the response of various parameters such as concrete
strength, percentage of steel on the behavior of RC beams. Column interaction
diagram and moment-curvature response of the beam and column sections have been
also studied. Finally, Chapter 6 draws the conclusion of the current research and puts
forward recommendations for future works for further development of the numerical
model.
6
Chapter 2
LITERATURE REVIEW
2.1 General
Reinforced Concrete (RC) inherently shows nonlinear behaviour due to its different
mechanical properties of constituent materials. RC structures show geometrical
nonlinearity also. A comprehensive knowledge of the behavior of structural concrete
elements under different types of loading is essential for producing safe, serviceable
and economic design of concrete structures. Traditionally structures are designed
following linear elastic analysis. However, to calculate deflections, effects of imposed
forces or second order effects the nonlinear behavior of RC has to be taken into
account. Growing architectural demands, combined with use of high strength concrete
and new construction methods, need accurate prediction of behavior of RC elements
or structures. In many cases, a usual linear elastic analysis no longer meets these
requirements and delivers large differences between the calculated and actual
deformations and internal forces. Analysis considering geometrical and material
nonlinearities has become popular with the availability of different types of
computers. The nonlinear analysis involves huge mathematical computations which
A brief review of previous studies on the application of the finite element method to
the analysis of reinforced concrete structures is presented in this section. A more
detailed description of the underlying theory and the application of the finite element
method to the analysis of linear and nonlinear reinforced concrete structures is
presented in the state of- the-art reports of the American Society of Civil Engineers in
One ofthe earliest publications on the application of the finite element method to the
.,If
7
analysis ofRC structures was presented by Ngo and Scordelis (1967). In their study,
simple beams were analyzed with a model in which concrete and reinforcing steel
were represented by constant strain triangular elements, and a special bond link
element was used to connect the steel to the concrete and describe the bond-slip
effect. A linear elastic analysis was performed on beams with predefined crack
patterns to determine principal stresses in concrete, stresses in steel reinforcement and
bond stresses.
Selna (1969) analyzed beams and frames made up of one-dimensional elements with
layered cross sections which accounted for progressive cracking and changing
materials properties through the depth of the cross section as a function of load and
time.
Nayak and Zienkiewicz (1972) used plane stress elements to study the behavior ofRC
frame. They conducted two dimensional stress studies which included both the tensile
cracking and the elasto-plastic behavior of concrete in compression using an initial
stress approach.
Nilson (1972) introduced nonlinear material properties for concrete and steel and a
nonlinear bond-slip relationship into the analysis and used an incremental load
method of nonlinear analysis. Cracking was accounted for by stopping the solution
when an element reached the tensile strength, and reloading incrementally after
redefining a new cracked structure. The method was applied to concentric and
eccentric reinforced concrete tensile members which were subjected to loads applied
at the end of the reinforcing bars and the results were compared with experimental
data.
For the analysis of RC beams with material and geometric nonlinearities Rajagopal
(1976) developed a layered rectangular plate element with axial and bending stiffness
in which concrete was treated as an orthotropic material. RC beam and slab problems
had also been treated by many other investigators such as Lin and Scordelis (1975),
.,
8
Bashur and Darwin (1978), Cook and Mitchell (1988), Vecchio (1989), Stevens et al.
(1991), Vecchio and Collins (1993), Polak and Blackwell (1998) etc.
There are two basically different approaches which have been used so far for the
analysis of RC elements by the finite element method. These are the modified
stiffness approach and the layered approach. The former is based on an average
moment-curvature relationship which reflects the various stages of material behavior,
while the latter subdivides the finite element into imaginary concrete and steel layers
with idealized stress-strain relations for concrete and reinforcing steel.
Lin and Scordelis (1975) utilized layered triangular finite elements in RC shell
analysis and included the coupling between membrane and bending effects, as well as
the tension stiffening effect of concrete between cracks in the model.
The behavior of RC columns under increasing load has been determined by many
researchers to simulate the behavior. One of the earliest publications in this regard
was by Pfrang and Siess (1964). Here, a study had been made on the effects of several
variables known to influence the behavior and capacity of the column. Concentrically
and eccentrically loaded unrestrained RC columns had been tested and studied at
some length. Since there was a relatively little information available concerning
eccentrically loaded column with end restraints, all analytical studies of this problem
had been directed towards the determination of the column capacity. To observe the
thorough behavior of the column that affects the strength its behavior under increasing
load was required. Therefore, the writers had presented an analytical method for
determining the behavior of the columns that was not dependent on the assumption of
a predetermined deflected shape for the column, and does not require extensive
simplifying assumptions as to the character of the structural system, or the cross-
sectional properties ofthe members which made up the system.
Most of the analysis methods for RC columns are based on equilibrium and
compatibility. The material properties of the section were incorporated to observe the
realistic behavior. The elastic analysis of a RC beam section can be carried out by
•
9
means of the transformed area concept, where steel areas in compression and tension
are replaced by equivalent concrete areas. Najmi and Tayem (1993) used the
transformed area to find the column behavior theoretically. They analyzed the short
RC columns to predict the strength. Since the column is short the geometric
nonlinearity effect was small enough. The failure load was obtained by using the
classical bending theory. In the design of slender RC columns, the ACI Building Code
(ACI 318-89) uses the moment magnifier design method to account for the second
order effects. The accuracy of this approach is strongly influenced by a realistic
determination of the effective flexural stiffness of the column, which depends to a
large extent on cracking, inelastic, and nonlinear behavior of concrete and loading
conditions. Therefore, considerable researches have been directed for the
development of nonlinear and inelastic analysis methods of RC columns. In spite of
the large number of previous studies on the nonlinear finite element analysis of
reinforced concrete structures, only few conclusions of general applicability have
been arrived at. Recognizing that many of the previously proposed models and
methods have not been fully verified so far, efforts for the development ofthe models
and methods have been going on around the world and it is the intent of this study to
contribute in this context.
The responses of RC ductile moment resisting frames subject to lateral loads are
being studied since the early 1960s. In ductile frames, the connection of beams to
columns must be capable of resisting large lateral loads caused by earthquake or wind.
The structures will undergo large deformations under such severe loading. It is
therefore desirable that the flexural capacity of the beams developed and maintained
under loads. This would cast aside the undesirable mode of failure caused by the
deformation of plastic hinges in the columns which may jeopardize the overall
stability of the frame. The bending moment, shear and axial forces, and deflections of
RC frames at any stages of loading from zero to ultimate load can be determined
analytically using the conditions of static loading equilibrium and geometric
•
compatibility, if the moment-curvature relationships of the sections are known.
However, difficulties are caused by the nonlinearity of the moment-curvature
relationships, and a step by step procedure, applying load increments, is generally ()
,.
••••
10
necessary. In addition, the moment-curvature relationship of sections carrymg
moment and axial force is dependent not only on the section geometry and material
properties but also on the level of axial force. This interdependence means that the
moment-curvature relationship for each section must be recomputed at each increment
ofloading.
Lazaro and Richards (1973) studied the behavior of RC frame under load by a
successive linear approximation method based on the stiffuess method of analysis
which can be used to predict the behaviour of the frame from zero to ultimate load. In
this method the members of the frame were divided along their length into small
elements. At each load level the flexural rigidity, corresponding to the particular
bending moment and axial force at each element, was obtained from the appropriate
point on the moment-curvature relationship. Members were assumed to be uncracked
for the initial load increments, and deformations are determined using the uncracked
section flexural rigidity. When the cracking moment was reached the flexural rigidity
of the element was recomputed on the basis of the cracked section. At higher loads,
when the stresses at the elements enter the inelastic range, the flexural rigidity of each
element was adjusted to the corresponding point of the moment-curvature curve
calculated for that moment and axial force level. Eventually with further increments,
plastic hinges spread throughout the frame, and the ultimate load is reached when a
mechanism forms and no further load can be carried. The analytical and experimental
load-deflection and moment-curvature results obtained from a hinged based portal
frame showed good agreement. From that it is evident that the full analytical approach
to the behavior of RC frames at all stages of loading is lengthy and can be
successfully undertaken only with the aid of a computer having large storage. Since
the full approach is an analytical procedure, it would require a trial and error solution
if used in design and hence the approach could hardly be regarded as suitable for
design. Nevertheless, when more computer programs are available, the approach may
become a powerful analytical tool for evaluating or checking the structural
performance over the full range of loading including behavior at service loads and at
. the ultimate load. Geometric changes in the frame under load may also be included in,f:
such programs to allow for the effect of deflections on the internal forces. This would
11
account for moment magnification due to column deflections indicating any
instability effects, as well.
,
Computer programs for the complete analysis of RC frames are under development in
many parts of the world. The structural behavior of the RC bare frame using FE
method was carried out by Mehrabi and Shing (1997). They analyzed the bare frame
as a part of masonry infilled frame. They used a smeared crack finite element
formulation to model the concrete available in commercial packages for analyzing RC
frames.
Johansson (2000) analyzed the RC frame comer using FE program DIANA. Here 2D
plane stress models were used to simulate the concrete. The cracking of the concrete
was modeled using the smeared crack approach with fixed cracks. The perfect bond
between the reinforcing bars and concrete was used to simulate the general response
of specimens. AI-Chaar et al. (2002) studied the bare frame as a part of infilled frame
to observe the effect of the infill on the response of the frame. Programs suitable for
general application are still awaited.
o
12
Chapter 3
DEVELOPMENT OF A NONLINEAR NUMERICAL MODEL - NLRCF
3.1 Introduction
The finite element method is a general method of structural analysis in which the
solution of a problem is approximated by the analysis of an assemblage of finite
elements which are interconnected at a finite number of nodal points. The finite
element method is now well accepted as the most powerful general technique for the
numerical solution of a variety of engineering problems. Applications range from the
stress analysis of solids to the solution of acoustical phenomena, neutron physics and
fluid dynamic problems. Indeed the finite element method is now established as a
general numerical method for the solution of partial differential equations subject to
known boundary and initial conditions.
For linear analysis the finite element method is now widely used as a design tool. A
similar acceptance in nonlinear analysis problems depends on two major factors. First,
the increase in computational effort which is required for nonlinear problems
necessitate that considerable computing power be available at low cost to the
designer. Developments in the last two decades have ensured that high-speed digital
computers have gradually become available to the average designer and indications
are that reductions in unit computing costs will continue at an accelerating pace.
The second major factor is related to the level of complexity of nonlinear analysis.
Before application of nonlinear methods can become commonplace in design
situations, the accuracy and reliability of the proposed models has to be established
beyond doubt. The developments of improved element characteristics and more
efficient nonlinear solution algorithms as well as the experience gained in their
application to practical engineering problems have ensured that nonlinear finite
element analysis can now be performed with some confidence.
14
Thus, barriers to the wide use of nonlinear finite element techniques are gradually
removed. Nevertheless, difficulties still abound whose solution will require much
III
Range I: Elastic
Range II: Cracking
Range III: Steel Yielding or
Concrete Crushing
I, ,
,,
,,
,,
DEFLECTION
The material properties of concrete and steel are also strain-rate dependent to a
different extent. Because of these differences in short- and long-term behavior of the
constituent materials, a general purpose model of the short- and long-term response of
RC members and structures should be based on separate material models for
reinforcing steel and concrete, which are then combined along with models of the
interaction between the two constituents to describe the behavior of the composite
reinforced concrete materia!. This is the approach adopted in this study. The
assumptions made in the description of material behavior are summarized below:
,- .•• +
16
In the following the behavior of each constituent material and the derivation of the
corresponding material stiffuess matrix are discussed separately. This is followed by
the model of the interaction between reinforcing steel and concrete through bond. The
superposition of the individual material stiffuess matrices to form the stiffuess of the
composite reinforced concrete material and the numerical implementation of this
approach in the nonlinear analysis of beams, columns and frames are discussed in the
next chapter.
3.2.2 Concrete
3.2.2.1 Behavior of Concrete
8 (401
••-e- &
1301
~
~ 4
••
t•• 3 (201
!
e 2
(101
~
1
Fig. 3.2: Stress-strain curves for concrete cylinders loaded in uniaxial compression
(Park & Paulay, 1975)
18
In describing the uniaxial stress-strain behavior of concrete many empirical formulas
have been proposed. These are summarized by ASCE (1982). Figure 3.3a presents the
simplest of the nonlinear models, the linearly elastic-perfectly plastic model, which
was used by Lin and Scordelis (1975) in a study of reinforced concrete slabs and
walls. Figure 3.3b presents the inelastic-perfectly plastic model proposed by the
European Concrete Committee (CEB 1978) made up of a parabola and a horizontal
line. The model proposed by Hognestad (1951) is presented in Figure 3.3c. This
model is capable of representing quite well the uniaxial stress-strain behavior of a
wide range of concretes. Finally, Figure 3.3d presents a piecewise linear model in
which the nonlinear stress-strain relation is approximated by a series of straight line
£ £
(al (ib)
(J
(e) (d)
segments. Although this is the most versatile model capable of representing a wide
range of stress-strain curves, its use is restricted to cases in which experimental data
for the uniaxial concrete stress-strain relation are available.
In the present study the model of Hognestad (1951) is used after some modifications.
These modifications are introduced in order to increase the computational efficiency
19
of the model. In typical reinforced concrete beams, columns and frames which are
subjected to bending, the maximum compressive stress at failure does not reach the
compressive strength. This means that the compressive stresses in most of the
member reach a small fraction of the compressive strength at failure. The behavior of
these members is, therefore, dominated by crack formation and propagation, and the
yielding of reinforcing steel.
The parabolic shape as presented in Figure 3.4 where f'c is the crushing (ultimate)
strength of concrete and Sm is the strain corresponding to f'c can approximate this
uniaxial stress-strain behavior of concrete. The inclusion of falling branch is
important and vital for the successful development of a numerical model to be able to
predict the ultimate limit state of collapse for RC structures and its elements. The
assumptions employed in the development of numerical model are based on the
observed behavior of concrete and can be summarized as nonlinear stress-strain
f~ --------------------
stress (f)
i
-----1.~ m
strain (s)
I.Ssm
The tensile strength of concrete, generally less than 20% of the compressive strength,
can be obtained directly from tension specimens. While concrete compressive
strength is primarily used in the design of RC structures, its tensile strength is also of
consequence in various connections. The condition under which cracks form and
propagate on the tension side of RC flexural members depends primarily on the
tensile strength. There are considerable experimental difficulties in determining the
true tensile strength of concrete. For many years, tensile strength has been measured
in terms of modulus of rupture, the flexural tensile stress at which a beam starts
cracking. The ACI Code contains the recommendation that the modulus of rupture be
taken to equal to 0.62";f'c, f'c in MPa (7.5";f'c, f'c in psi) for normal weight concrete.
Since this study is more concerned with the flexural behavior of concrete structures, it
is felt that the contribution from tensile strength to overall flexural strength would be
small but can be incorporated in the model.
Additional cracks can form between the initial cracks, if the tensile stress exceeds the
concrete tensile strength between previously formed cracks. The final cracking state is
reached when a tensile force of sufficient magnitude to form an additional crack
between two existing cracks can no longer be transferred by bond from steel to
concrete. Figures 3.5c, 3.5d and 3.5e present the idealized distribution between cracks l)
•
21
of the bond stress, concrete tensile stress and steel tensile stress, respectively. Because
concrete is carrying some tension between the cracks, the flexural rigidity is clearly
I I
I I
I I
I (e) I
I I
:::J.c1,. ~
: ten I
I :Ii =t I
---:J
I (,) I.
I III
-J.-- I
(a) Portion of a beam
(b) Bending moment distribution
(c) Bond stress distribution
(d) Concrete tensile stress distribution
(e) Steel tensile stress distribution
(f) Flexural stiffness distribution in elastic range
greater between the cracks than at the cracks, as presented in Figure 3.5f (Park and
Paulay, 1975).
As the concrete reaches its tensile strength, primary cracks form. The extents of
cracks are controlled by the placement of the reinforcing steel. At the primary cracks
,
22
the concrete stress drops to zero and the steel carries the entire tensile force. The
concrete between the cracks, however, still carries some tensile stress, which
decreases with increasing load magnitude. This drop in concrete tensile stress with
increasing load is associated with the breakdown of bond between reinforcing steel
and concrete. At this stage a secondary system of internal cracks, called bond cracks,
Since cracking is the major source of material nonlinearity in the serviceability range
of RC structures, realistic cracking models need to be developed in order to accurately
predict the load-deformation and moment-curvature behavior ofRC members.
Typical stress-strain curves for reinforcing steel bars used in concrete construction are
obtained from the test of bars loaded monotonically in tension. For all practical
purposes steel exhibits the same stress-strain curve in compression as in tension.
120
•••
./
V - - (800)
-
..•
E
100
- ~
/ - (600)
.E 80
z
..•c
~ 60
/ - (400)
0.
:;; /'
~.40
t~ - (200)
Vi •••
20
Fig. 3.6: Typical stress-strain curves for steel reinforcement (Park & Paulay, 1975)
The modulus of elasticity of the steel is given by the slope of the linear elastic portion
of the curve. The modulus of elasticity of steel reinforcement is generally considered
5
as 2 x 10 MPa (29 x 103 ksi). The stress at yield point, referred to as yield strength, is
. a very important property of steel reinforcement for RC structures.
For the analysis and design of RC structures it is required to define the shape of
stress-strain behaviour of steel reinforcement. There are a number of stress-strain
relationships available to idealize the steel as presented in Figure 3.7. The first
idealization neglects the strength increase due to strain hardening and the reinforcing
steel is modeled as a linear, perfectly plastic material, as presented in Figure 3.7a.
This assumption underlies the design equations of the ACI code (Park and Paulay,
1975). If the strain at the onset of strain hardening is much larger than the yield strain,
this approximation yields very satisfactory results. This is the case for low-carbon
steels with low yield strength. If the steel hardens soon after the onset of yielding, this
approximation underestimates the steel stress at high strains. In several instances it is
necessary to evaluate the steel stress at strains higher than yield to more accurately
assess the strength of the member at large deformations. This is, particularly, true in
seismic design, where assessing the available ductility of a member requires that the
behavior be investigated under strains greater than yield strain of steel. In this case
more accurate idealizations which account for the strain hardening effect are required,
as presented in Figures 3.7b and 3.7c. The parameters of these models are the stress
and strain at the onset of yielding, the strain at the onset of strain hardening and the
stress and strain at ultimate. These parameters can be derived from experimentally
obtained stress-strain relations.
I
24
r.
B Horizontal C
f,
tan' -£.
" "
'.
(,rl
r.
D
B C
f,
I
I
I
tan8aE.1
6 I
I
"" '.
'"
(hI
f,
f•• -------------- I
I)
I
I
I
B C I
f, ,
I
tanO-E.
,,
I
6 , I
'.
"" ' .. ..
'
1,1
( C) Complete curve
Fig. 3.7: Idealizations for the stress-strain curve for steel in tension or compression
In this study the reinforcing steel is modeled as a linear elastic, linear strain hardening
material with yield strength fy as presented in Figure 3.8. The reasons for this
approximation are: (I) the computational convenience of the model; (2) the behavior
of RC members is greatly affected by the yielding of reinforcing steel when the
structure is subjected to monotonic bending moments. The elastic strain hardening
model of steel reinforcement has been successfully used in many analyses of RC
structures (Ngo and Scordelis, 1967; Bashur and Darwin, 1978). The stress-strain
curve of reinforcing steel is modeled by a bilinear curve in tension and compression.
cr
E'2
fy ------
04
-cu
c
Cy Cu
Bond is the interaction between reinforcing steel and surrounding concrete. The force
transfer from steel to concrete can be attributed to three different phenomena: (I)
chemical adhesion between mortar paste and bar surface; (2) friction and wedging
action of small dislodged sand particles between the bar and the surrounding concrete;
and (3) mechanical interaction between concrete and steel. Bond of plain bars derives
26
primarily from the first two mechanisms, even though there is some mechanical
interlocking caused by the roughness of the bar surface. Deformed bars have better
bond than plain bars, because most of the steel force is transferred through the lugs to
concrete. Friction and chemical adhesion forces are not negligible, but secondary and
tend to decrease as the reinforcing bars start to slip.
Since bond stresses in reinforced concrete members arise from the change in the steel
force along the length, the effect of bond becomes more pronounced at end
anchorages of reinforcing bars and in the vicinity of cracks. In the simplified analysis
of RC structures complete compatibility of strains between concrete and steel is
usually assumed, which implies perfect bond.
In nonlinear analysis of reinforced concrete structural elements, the flexural and axial
rigidities of the members are function of current state of strain and/or loads. This, in
general makes it very difficult to obtain a closed form solution for evaluating the
sectional stiffuess. It is therefore required that the adopted numerical procedure
should evaluate these stiffuesses at different load levels. This model should
incorporate the behavior at different stages of loading such as cracking, yielding, and
ultimate and their respective curvatures of the structural elements. A sudden change in
the slope or stiffuess may introduce numerical difficulty in the model. Basic
assumptions used in the development of numerical model -NLRCF (Nonlinear
numerical model for analyzing Reinforced Concrete Frame) are as follows:
,\'ttl ..•...
-...;
27
The finite element analysis of a continuum starts with the subdivision of the physical
system into an assemblage of finite elements. The axial rigidity and flexural rigidity
can be defined as:
Sa = P/sa , and
Sf = M/<p
where,
Sa = axial rigidity
Sf = flexural rigidity
P = axial force acting at the section centroid
M = bending moment applied to the section
s.= axial strain of the section due to axial load measured at the section centroid
<p= curvature of the section
To calculate Sa and Sf due to load (PI, Ml) the corresponding sectional defomlation
vector (SI, <PI)are calculated first. If the sectional deformation due to an initial load
vector (Po, Mo) is known as (so, <Po)the following relations can be written:
~ Po M
= + (3.1 )
M\ Mo f\.M
8\ 80 !!.8
+ (3.2)
((1\ ((10 !!.((1
From (3.1) and (3.2) it is obvious that the incremental deformation (!!.s, !!.<p)
corresponds to the incremental load (!!.P, !!.M).
28
The relationship between an infinitesimal load and an infinitesimal deformation can
be formulated as shown in matrix form
where,
K, is the sectional stiffness matrix. Using the infinitesimal changes in the load and
deformation vector by finite (incremental) changes (3.3) can be written as
•
M =[K]!'>.& (3.4)
!'>.M '!'>.rp
=[r',] M (3.5)
!'>.M
Equation 3.5' represents the solution algorithm for the incremental deformation.
Usually, more than one iterative cycle is necessary for the solution to converge.
Using the current state of strain profile under the current forces P and M, the sectional
forces such as axial force, shear force and bending moment are obtained.
29
3.4 Stiffness of a Beam Element
The finite element method is familiarized in the field of civil engineering through
attempts to generalize the stifthess method to deal with structural continua. The
method has a number of features that facilitate automatic handling. Local stiffuess
matrix of members (or elements) and joint loading from members can be determined
without considering the overall structural configuration. The overall stiffuess matrix
of a structure can be assembled automatically from the local stiffness matrices of
members. The overall stiffness matrix can be solved to determine joint forces and
displacements by using standard solution procedures.
Stiffness of an element relates the forces exerted on the member ends to the
displacements of the ends. The stiffuess matrix for a beam element as presented in
Figure 3.9 with three degrees of freedom at each node can be written as [F]=[k][il]
where [k] is the stiffness matrix that is well known in the structural theory. The
stiffuess matrix remains valid if the force-displacement relationship is linear. For a
structure behaving in a nonlinear manner linear relationship no longer holds true. In
other words, it is not possible to express stifthess matrix explicitly to represent the
force-displacement relationship in a linear form. To overcome this problem, the
incremental/iterative method of analysis is used. In the incremental approach the load
is applied in small increments and it is assumed that the structure behaves linearly in
each increment. Thus for each increment the linear force displacement relationship
can be applied but it is different for each different increment. So the incremental
force-displacement relationship for an element can be written as: [oF] = [k][oil] where
[k] can be defined as instantaneous stiffuess matrix which corresponds to a particular
state of the element. At the beginning of an increment the nodal forces and
displacements and hence strain and curvature throughout the length of the elements
are known, which correspond to the total applied load up to the previous increment.
Representing the displacement pattern for the current increment of load by a suitable
shape function, the increment in curvature and strain throughout the element can be
obtained. Using the incremental values at the nodes, the corresponding increment in
nodal forces can be obtained from basic stiffness properties of a RC section.
30
Since the behavior of a RC section is nonlinear the stiffness becomes a function of not
only modulus of elasticity and area but also of the state of strain and hence stress and
stiffness properties vary along the length of members. This is due to the fact that
strain distribution varies over the length of a member. However, using a large number
of small elements the variations of stiffness properties over the length of member can
Sign convention
I•• L
(3.7)
[JF] = [KJJLi]
where, [K,) is the stiffness matrix of the element, SII, SIZ, SZI and Szz are the stiffness
terms corresponding to the sectional properties and state of strain at the middle of the
""
31
element ( Rouf, 1984). As the derivation is based on the properties of mid-section of
the element, the whole element is to be considered cracked if the mid-section is
cracked; otherwise it is to be considered as uncracked. The stiffuess matrix for an
uncracked is presented in Table 3.1 and for a cracked element in Table 3.2.
6£0 01
--~~a
L2
1
4£0 0
---a
L
x,
B
I I
, I
TO E57
I. L/2
+
X
L/2 -I Section
-kx-x
Strain at
x,
B SI
I I
, I
TO }V
I- L/2
'I-X L/2----1 Section Strain at x-x
where,
Ao = sectional area of whole section (BT)
10 = second moment of area of whole section with respect to centroidal plane of
whole section (= BT3/12)
Eo = initial tangent modulus ( = 2ic/sm)
X = 3DE2 - 2D2E3
Y = D3E4 - 2D2E3 + 3DE2/2
Ej = i - SI/Sm for i = 2, 3, 4
D = depth ratio
33
The contribution of stiffness matrix [Ks2] for RC sections due to reinforcing steel is
derived by Rashid (1988) and can be written as
o o E,A,(d_L)
E,A, _E,A, (d-L) J~A L
L 17£,A, (d-L)' L L _11£,A, (d-L)' 6E,A,(d-L)'
o E 6E,A, (d-L)' 0 E E
L'
_ E,A, (d-L) 6E,A, (d-L)' 4E,A, ,E,A, (d-L) _ 6E,A, (d-L)' 2E,A,(d-L)'
L 13 --(d-L) L 13 L
- -----------------------------L---------- T---------------------------------------- --
_ E,A, 0 E,A, (d _ L): E,A, 0 _ E,A, (d-L)
L -17£,A, (d ,L : L 11£,A, (d L)' L
o E ~ -L) _ 6E',A,(d _ L)' : 0 ~ - _ 6E,A,(d-L)'
L, "
EA, 6EA, , EA, 6EA, 1..-
-'-(d-L) -'-(d-L)' 2E,A, (d-L)' :--'-(d-L) --'-(d-L)' 4E,A, ,
L 13 L ! L 13 T(d-L)
where,
Es = the slope of the appropriate part of the stress-strain curve as shown in
Figure 3.8 depending upon the strain present in steel
As = area of the steel
d = depth ofthe steel from the top of the section
L = total depth ofthe section/2
Here, it is noted that the stiffuess terms for reinforcing steel appear in linear form,
due to the fact that the basic stress-strain characteristics of steel is a combination of
two linear parts as shown in Figure 3.8. The expressions presented here are related to
the first part of the curve representing linear elastic stress-strain relationship for steel
(before yielding is started). The stiffuess terms corresponding to other part of the
stress-strain curve (plastic or strain hardening range) can easily be obtained from the
expressions presented by just replacing the term E" with the slope of the
corresponding portion of the curve. Total stiffuess [K] of a RC section is the sum of
stiffnesses due to stiffuess of concrete [Ks1] and steel reinforcement [Ks2], respectively
as presented follows
\.-
, ~
...•..
•.' 1 -
34
can be written as
[oF] = [K][M] (3.9)
The algorithm used in this model, the nonlinear constitutive relations is satisfied by
iterative successive corrections. In each iteration the forces are determined from the
corresponding strain increments at the end of the previous iteration step, while the
material properties are only updated at the beginning ofthe next load step. Since in a
nonlinear problem the forces, which are determined during the iterative phase of the
algorithm are not, generally, in equilibrium with the applied loads, unbalanced nodal
forces result. These are corrected during the subsequent iteration until a specified
tolerance is satisfied. The unbalanced nodal forces are the difference between applied
and resisting or equivalent forces as presented below
The resisting nodal forces are statically equivalent to forces, which results from the
current deformation state of the element. The nodal forces thus obtained for all
elements are summed up to get the equivalent load vector for the structure. Details of
the numerical implementation along with a summary of the algorithm and a
discussion of the convergence criterion are presented in the following section.
The numerical implementation of the finite element model requires the solution of the
equation 3.9. This is a system of simultaneous nonlinear equations, since the stiffness
matrix [K] in general, depends on the displacement vector {o}. The solution of this
system of nonlinear equations is typically accomplished with an iterative method. The
load vector {F} is subdivided into a number of sufficiently small load increments,
which are successively applied (Fig. 3.10) where r is the corresponding displacement.
t
• .'
35
At each load step a linear approximation of the stiffuess matrix [K] is established and
the resulting system of linear equilibrium equations is solved for the displacement
increments which correspond to the applied load increments. Since the stiffness
matrix [K] changes under these displacement increments, the resisting forces of the
structure do not equilibrate the applied loads and unbalanced loads result as presented
in equation 3.10. In the subsequent correction phase the displacement increments are
iteratively improved, until a specified convergence criterion is satisfied. If no
correction phase is included in the nonlinear analysis algorithm, the numerical error
grows from one load step to the next and the numerical solution drifts away from the
------
.-
./'
"
~ IACiual1
. ~ ~ - - Numerical
Depending on how the stiffuess matrix [K] is updated during the correction phase, the
iterative method can be classified into three broad categories even though variations
of these schemes are also possible: the initial or constant stiffness method, the tangent
stiffness method and the secant stiffuess method (Taylor and Hinton, 1980).
36
The tangent stiffness method requires the smallest number of iterations to arrive at the
solution, but has the disadvantage that the stiffness matrix [K] needs to be reformed
and triangularized at each iteration. The initial stiffness method, on the other extreme,
requires the largest number of iterations, but the stiffness matrix [K] is only formed
and triangularized once at the beginning of
the load step. Generally, the tangent
stiffness method is more efficient than the
initial stiffness method for models with a
small number of degrees of freedom, even
though the selection of one method over
another often depends on numerical
stability considerations.
lal
Even though it is possible to use any
stiffness matrix in the first iteration of a
new load increment, which constitutes the
advancing phase of the solution algorithm,
it is quite common to use the tangent
If stiffness matrix for this purpose. The
nonlinear solution scheme selected in this
study is the modified N ewton- Raphson
method that uses the initial stiffness matrix
at the beginning of the load step in
combination with the same stiffness matrix
during the subsequent corrective iterations,
•
(a) Initial Tangent Method
(b) Tangent Stiffness Method
(c) Secant Stiffness Method
Fig. 3.11: Iterative Methods
37
Every nonlinear analysis algorithm consists of four basic steps: the formation of the
current stiffness matrix, the solution ofthe equilibrium equations for the displacement
increments, the determination of status of strains and internal forces in all the
elements of the model and the convergence check. These steps are presented in some
detail in the flow diagram in Figure 3.13.
F,
F,
li----IAdUB1wl
. Numerical.
F,
o r
This section outlines a step by step procedure for the nonlinear solution algorithm
and, the corresponding operations are presented below and are also presented in
Figure 3.13. In this development, the sectional state of strains under different load
level is determined. In other words, the deformation vector (el, <PI)under a load vector
(PI, Md is to be determined if the section response (eo, <Po)under initial load vector
(Po, Mo) is known.
.',,.,
....;..\4
.( :
..•.... .
,
"'I
38
on
E E Determine the element strains for the current displacements
'Ei" Ei'"
~
"
.S
'"
"'~"
:;;
"'o" s 0 Use the material constitutive model to determine
..:2
=a "- the internal displacements and internal nodal forces
~
0
"- Determine the difference between external and
internal nodal forces
No
Yes
Print displacements, strains and forces
No
The initial load vector for the first load increment will be of course (0, 0). In the
following steps, the procedure for obtaining the deformation vector (EI, qJI) is
summarized.
1. Assume an incremental load and determine load vector from Eq. 3.1
to
2. Calculate the sectional stiffuess matrix from Eq. 3.8 for a RC section. ,r,
••
39
3. Determine the incremental deformation vector (t.e, t.<p) from Eq. 3.5
4. Calculate the sectional deformation vector (el, <PI) from Eq. 3.2
5. Calculate the load vector corresponding to the deformation vector obtained from
Step 4. Compare the load vector so obtained with the load vector (PI, MI).
If the difference between two load vectors is within an acceptable tolerance, the
solution for this load increment would be the deformation vector obtained from Step
4. Otherwise, the difference between load vector (PI, MI) and load vector calculated
in this step should be used to obtain a more accurate value for deformation vector by
using this difference vector as an incremental load vector and repeating Steps 3
through 5.
The criterion for measuring the convergence of the iterative solution is based on the
accuracy of satisfying the global equilibrium equations or on the accuracy of
determining the total displacements. The accuracy of satisfying the global equilibrium
equations is controlled by the magnitude of the unbalanced nodal forces. The
accuracy of the node displacements depends on the magnitude of the additional
displacement increment after each iterative cycle. The mixed convergence criterion is
used in this study. The displacements obtained at the end of each iteration are checked
against the total displacements obtained up to that iteration in the increment.
8N
_,_< c
- VD
8,
where, OiN = displacements obtained in the nth iteration in the ith increment ofloading
Oi = total displacements obtained up to the nth iteration in the ith increment
The second convergence test is carried out on the pseudo load vector. The elements of
40
pseudo load vector (Rouf, 1984) and the applied load vector of the increment are
normalized to scalar values by
NDF
A= LF/
j=!
To test the convergence, both these displacement and force tolerances are applied. In
the present study tolerance has been used I % both for displacement and load
convergences. Any number of iterations and any limit of tolerance for both the
displacement and force can be used in this model. The failure load for a structure is
assumed to occur at a load level for which a large number of iterations are required
for convergence. This will occur when the structure is approaching collapse. The
displacement convergence test does not, however, cause the program to terminate the
execution.
In the nonlinear analysis of RC structures the load step size must be small enough so
that unrealistic "numerical cracking" does not take place. These spurious cracks can
artificially alter the load transfer path within the structure and result in incorrect
modes of failure. Crisfield (1982) has shown that such numerical disturbance of the
load transfer path after initiation of cracking can give rise to alternative equilibrium
states and, hence, lead to false ultimate strength predictions.
41
Mesh density is important in any analysis of a structure by using finite element. Since
increasing number of elements increases the computational time researchers have
been satisfied with some acceptable error. In the present model beam element has
been used. Therefore, the meshing of the member consists of length wise
discretization of the member. There are many approaches to select the mesh density
and corresponding error. The present model is based on the slope-deflection method
of an element. Thus an easiest approach is to compare the outputs such as deflection
at a particular node in the member. In this study a number of analysis for RC beams,
columns and frames have been carried out. For example a sensitivity analysis for the
RC beam A-3 (Bresler and Scordelis, 1963) is presented here to obtain optimum
number of elements. This beam had a cross-section 552 x 305 mm and reinforced
with 6-# 9 bars as tension reinforcement and 2-#4 bar as compression reinforcements.
The effective span length of the beam was 6.40 m and was subjected to single point
loading at midspan. The deflection obtained at the centre of the span of the member
by the model is plotted against the number of elements and is presented in Figure
3.14.
I
li~-------------""""&\ 11:0 H
I- 6.4 m
'I 305 rom
40
38
S
536
8 ~~-----
'B
d5 34
a"
32
30
200 250
o 50 100 150
No. of elements
Fig. 3.14: Deflection at midspan of the RC beam A-3 (Bresler & Scordelis, 1963) as a
body.
4.1 Introduction
Over the past fifty years many researchers have carried out a number of experimental.
and analytical investigations and important advances have been made for RC beams,
columns and frames. A number of correlation studies have been conducted which are
presented in this chapter with the aim of establishing the capability of the proposed
model to simulate the load-deflection and moment-curvature responses ofRC beams.
The classic series of beam tests conducted by Bresler and Scordelis (1963) some 42
years ago to investigate RC members in shear, is commonly regarded as a benchmark
against which finite element analysis model can be calibrated. The use of Bresler-
Scordelis beams as a benchmark series is due to, in part, to the high quality and
thorough documentation of the tests. The 12 beams tested by Bresler and Scordelis
consisted of four series each containing three beams: each series differ in amount of
reinforcement, span length, cross sectional dimension, and concrete strength. Al1 the
beams were of rectangular cross section with same overal1 depth of 552 mm. The
bottom longitudinal bars were provided by NO.9 bars while NO.4 bars were used for
al1 the top longitudinal bars. The beams were characterized by three different modes
of failure: diagonal-tension, shear-compression and flexure-compression. Three
beams designated as A-3, B-3 and C-3 by Bresler and Scordelis (1963) among 12
failed due to flexure-compression. These beams had web reinforcement and greater
span length. The shear span ratio for these beams had a nominal value of 7. The
beams failed by crushing of the compression zone near midspan at the section of
maximum moment. Cross sectional details and material properties of beams A-3, B-3
and C-3 are provided in Table 4.1 and Table 4.2, respectively and the steel positions
are detailed in Figure 4.1. These three beams had span length of 6.40 m and subjected
44
to a single point loading at midspan. These three flexure critical beams are analyzed
with the model NLRCF and compared with the test results. Figure 4.2 presents load-
deflection and Figure 4.3 presents moment-curvature responses of these beams. Table
4.3 presents the comparison between the test and numerical results.
374mm
0
63.5 mm
_ 63.5 mm
Table 4.1: Summary of cross sectional details of beams (Bresler & Scordelis, 1963)
Spacing # 2
Beam No. Bottom steel p (%) Top steel pi (%)
stirrups(inch )
6 -#9 2.73 2 - #4 0.182 81/4
A-3
5-#9 3.06 2 - #4 0.245 7'/z
B-3
4-#9 3.63 2 - #4 0.363 81/4
C-3
Table 4.2: Material properties of Bresler-Scordelis beams (Bresler & Scordelis, 1963)
A. Reinforcement
B. Concrete
500
400
~ 300
~
~
.",
"
0
>-l
200
100
--A-3model ....... A-3Test
--B-3 model - - - - B-3 Test
----.--C-3 model _. -. - C-3 Test
o
o 5 10 15 20 25 30 35 40
Deflection (nun)
1000
800
_---0
••••.•• [>
E>
o
o 0.002 0.004 0.006 0.008 0.01
Curvature (rad)
A-3 468 476 459 0.98 1.02 35.8 39.7 35.1 0.90 1.02
B-3 356 401 385 0.91 0.92 35.3 40.7 35.3 0.87 1.00
C-3 270 275 270 0.98 1.00 36.8 37.7 34.3 0.98 1.07
From Figure 4.2 and Table 4.3 it is observed that a good correlation exists between
test and numerical results. All these three beams failed due to crushing of the
compression zone near midspan at the section of maximum moment as test results
indicated by Bresler and Scordelis (1963) and also by the model. Since the modulus
of elasticity of the concrete was not reported by Bresler and Scordelis (1963), two
different Ee are used in the analysis and the obtained load, deflection results are
compared in Table 4.3, while the moment-curvature response is compared as
presented in Figure 4.3:
Bangladesh has scarcity of natural aggregate like stone for making RC concrete
structures. So it makes people to depend on artificial aggregate like crushed brick. An
investigation by Akhtaruzzaman and Hasnat (1986) had presented that with brick
aggregate, high quality concrete can be easily produced and its physical and
mechanical properties compares favorably with those of normal weight concrete. The
3
lower unit weight of brick aggregate concrete, which varied from 2000 to 2080 kg/m
gives an added advantage. Forty eight rectangular beams with identical reinforcement
and cross section were tested in shear and flexure under two-point loading. All these
beams had effective span length of 2.21m and two NO.6 bars (yield strength of 397
MPa and modulus of elasticity of201.5 x 103 MPa) were used as tensile reinforcement
for the beams. The coarse aggregate consisted of 19 mm graded, well burnt crushed
<-
•
•.1
47
surface clay bricks and sand was used as fine aggregate. Type-I Portland cement was
used in all cases. The variables were concrete strength and shear span to effective
depth ratio (aid). The different modes of failure observed during test were deep-beam
type splitting failure, shear-compression failure, diagonal-tension failure, teeth-action
failure, and typical flexural failure. Among 48 beams, beams designated by the
authors A6-I, A6-Il, BrI, C7-1 and Ds-I failed in flexure. These beams had higher
shear-span ratio compared to other beams in that series. Only these beams are
considered here for analysis by NLRCF and are compared with the numerical results
and ACI Code (Akhtaruzzaman and Hasnat, 1986) for normal weight aggregate. The
cross-sectional details and material properties are presented in Table 4.4.
The modulus of elasticity of concrete used in the analysis was Ec = 33 w/5 ;frc, psi as
suggested by ACI Committee 3 I 8-7 I and unit weight of brick aggregate was
considered BOpcf.
The load-deflection responses of the beams are presented in Figures 4.4 and 4.5 and
are compared with the results obtained by NLRCF. The deflections refer to mid span
deflection due to applied load only. The theoretical midspan deflections were
computed for service load; ultimate applied load/I. 7, using the 1983 ACI building
Code equations. A comparison has been made in Table 4.5 between the test and
obtained load by NLRCF and among the computed, analytical and measured
experimental short-term deflections.
" ..
48
80
n
.' t>
60
.' t>'
~
~
:;'40
8
..-l
20 -
•.• 8- .• A 6-11test • A6-II model
0
0 2 4 6 8 10 12
Deflection (mm)
Fig. 4.4: Load-midspan deflection response ofthe beams A6-II and B7-I
(Akhtaruzzaman & Hasnat, 1986)
80
.8
60 -
20 -
•.• 8- .• C7-1test __ C7-Imodel
o
o 2 4 6 8 10
Deflection (mm)
Fig. 4.5: Load - midspan deflection response of the beams C7-I and D8-I
(Akhtaruzzaman & Hasnat, 1986)
49
Table 4.5: Experimental and predicted ultimate load and deflection at midspan
,Akhtaruzzaman & Hasnat, 1986)
Ultimate load Deflection at
Beam aid (kN) P u,testlP u,pred. Pu/1.7 (mm) 8testJ8prcd. .6.AcII6.pred.
Pu test Pu red. 6.test 8ACI 6. red.
A,-II 4.0 74.2 72.2 1.03 4.95 4.34 4.29 1.15 1.01
B,-I 4.5 60.8 62.7 0.97 4.45 4.09 4.00 1.11 1.02
C,- I 4.5 64.7 62.2 1.04 5.33 4.52 4.43 1.20 1.02
D, - I 5.0 54.2 52.2 1.04 4.19 4.45 4.53 0.92 0.98
From Figures 4.4 and 4.5, it is observed that the behavior of beams generally presents
a good correspondence with the test results. The deflection obtained by the model is
lower than the test value.
Two profiled steel-concrete beams and two reinforced concrete beams were tested by
Uy and Bradford (1995) under service loads. The four beams were tested to failure in
flexure using a two point loading arrangement that produces constant moment region
for a simply supported beam having span of 6.0m. The beams were designed to fail in
flexure rather than shear using the aforementioned loading arrangement, since the
flexural response was being considered for the ductility. Two RC beams among the
four beams had the same cross-sections and reinforcement details. The test results
showed that these two beams had the similar load-deflection response and moment
carrying capacities. Therefore, for comparison, one of the RC beams is considered. A
mean compressive strength of 42.6 MPa with the elastic modulus of concrete of
33300 MPa was obtained and used in the analysis. Deformed reinforcing bars were
used and tested in tension for the determination of the elastic modulus and tensile
strength. Y 12 and Y20 bars were used as longitudinal reinforcement in RC beams.
2
The elastic modulus of both the YI2 (area = 113 mm2) and Y20 (area = 314 mm )
bars was 200,000 MPa. The yield strengths were 435 MPa and 439 MPa for the YI2
and Y20 bars, respectively. The cross-sectional details and arrangement of loading is
presented in Figure 4.6. The load-deflection and moment-curvature responses of the
RC beam and analytical results obtained using NLRCF are presented in Figures 4.7
, ...
50
and 4.8, respectively.
p p
2m 2m
I'
2m
0\- 0\- -I
I- 265 mm -I ~
40mm
E
T
E
0
0
"<t
I.50mm
45mm
T
Fig. 4.6: Cross-sectional detail and loading arrangement of the RC beam
(Uy and Bradford, 1995)
250
~~~
~---- ---------------
200 "
150
100
50 - - - -Test
--e- Predicted by NLRCF without tensile strength of concrete
. _-e ... Predicted by NLRCF with tensile strength ofconcrete(4.065 MPa)
o
100 125
o 25 50 75
.Deflection <rom)
Fig. 4.7: Load-deflection response of the RC beam (Uy & Bradford, 1995)
51
250
-----
--------- ---------
200
~150
~
""
~IOO
----Test
50
--G-- Predicted by NlRCF without tensile strength of concrete
...• - .. Predicted by NLRCF with tensile strength ofconcrete(4.065 MP.)
o
o 10 20 30 40 50
6
Curvature (radlmm) x 10-
Fig. 4.8: Moment-curvature response of the RC beam (Uy & Bradford, 1995)
250
200
50
__________Predicted by NLRCF without tensile strength of concrete
. _.e ... Predicted by NLRCF with tensile strength of concrete (4.065 MPa)
o
o 0.001 0.002 0.003
Concrete compressivestrain (mm1mm)
Fig. 4.9: Effect of tensile strength on the maximum concrete compressive strain of the
RC beam obtained by the model
52
From Figure 4.7 it is observed that there is a good correlation between the test and
numerical results. The measured deflections after yielding of the reinforcement
present a fairly large plateau indicating ductile failure as expected for an under-
reinforced beam and this is clearly obtained from analytical solution too. The collapse
load obtained by the model is 210 kN in both cases whether tensile strength of
concrete is considered or not. From the Figure 4.7, it is observed that the tensile
strength of concrete has insignificant effect on the collapse load. The deflection
obtained by the model agrees well with the test deflection. In Figure 4.9 a comparison
is presented for the moment corresponding to concrete compressive strain at each load
level considering tensile strength of concrete and ignoring it. From this figure it is
observed that the tensile strength has no significant effect on the maximum concrete
compressive strain. The tensile strength of concrete has significant effect on
deflection, curvature and concrete compressive strain at or near cracking load which
is observed from the above ,figures.
Test results indicate that brittle type of failure in over-reinforced beam can be
controlled by introducing ties or discrete steel fibers in the compression zone. For
such a beam, both ultimate ,strength and ductility can be enhanced by increasing
concrete strength. The stress-strain curves for concrete in compression obtained from
flexural tests are remarkably similar to those generated from uni-axially loaded
specimen. The analysis is based on flexural theory, but the use of stress-strain curves
for axially loaded specimen gives close predictions of the experiment data on
moment -curvature relationship and ultimate moment capacity of the beams. The
stress-strain curve used for normal strength concrete is also used here to describe the
behavior of high strength plain concrete. The test program had been carried out to
observe the flexural behaviour of over-reinforced HSC beams with and without
confinement in the compression zone by Mansur et al. (1997). Concrete strengths
ranging from 60 to 110 MPa were used to observe the flexural response of the beams.
The study included the effect of concrete compressive strength on the beam behavior.
The beams were designated as A4-0.0C, B4-0.0C, C4-0.0C, and D4-0.0C. It was
53
observed that failure of the beams was characterized by sudden crushing of concrete
prior to yielding of the steel.
Table 4.6: Comparison between test and predicted values of ultimate moment
(Mansur et aI., 1997)
Table 4.7: Comparison between test and predicted values of ultimate load and
deflection (Mansur et aI., 1997)
••••
54
175
105.3 MP
150
89.1
125 72.9
57.2
E
~ 100
5
~ 75
50
25
120
80
a
~
~
E 60 2%
~
::E
40
20
0
0 0.01 0.02 0.03 '0.04 0.05
Curvature(radlm)
A theoretical investigation is carried out for the beam A4-0.0C to observe the effect of
percentage of steel on the response of various parameters such as load, deformation,
stiffuess (force required per unit deflection of the member at centre of the span) and
curvature. Figures 4.12, 4.13, 4.14 and 4.15 present the variations of these parameters
due to percentage of steel.
200
30
~
"il150 i~
.3
•.. .g" 20
u
.~ 100 <Ij
•••
:5 0
10
50
o 0
o 2 4 6 8 0 2 4 6 8
% of steel % of steel
Fig. 4.12: Effect of% of steel on load Fig. 4.13: Effect of % of steel
on deflection at ultimate load
56
8 0.04
~
I ~
'" 0.03
<::.
~6
~
~ 0.02
t
u"
2 0.01
o 0
o 2 4 6 8 0 2 4 6 8
% of steel % of steel
Fig. 4.14: Effect of% of steel on stiffness Fig. 4.15: Effect of % of steel on
Curvature
From the above figures it is observed that an increase in tension steel increases the
ultimate load carrying capacity and stiffness but decreases the deflection and
curvature of the section.
The effect of reinforcing bar deformation pattern on flexural ductility was observed
by Tholen and Darwin (1998) through simply supported beam tests. The main test
parameters were relative rib area and reinforcement ratio. The specimens consisted of
two matched pairs of beams. One set of specimens contained two 8 No. bars,
providing a reinforcement ratio of 43% of balanced steel ratio based on material
properties. The second set of specimens contained 3 NO.8 bars with p = 0.68Pb. One
beam in each group contained conventional reinforcement, while the other beam
contained high relative rib area bars. The beams were 4.9 m long with nominal width
and depth of 305 and 406 mm, respectively. The specimens were tested as simply
supported beams with a concentrated load at the centre of the span. The nominal
bottom and side covers were 51 mm, No.3 closed stirrup with a spacing 152mm were
57
used to provide shear strength throughout the member, and NO.4 bars were used as
top reinforcement. All the bars met the requirements of ASTM A615. Grade 60 bars
were used for stirrups and top reinforcement. The yield strength of No. 8 bars was 538
MPa. Concrete compressive strengths were 36 and 33 MPa for the specimens with p =
0.43 Pb and 0.68pb. respectively. •
From the test results it was observed that the relative rib area of reinforcing bars did
not affect the displacement or rotational capacity of the beams in which plastic hinges
develop. As expected the specimens with three bars produced higher load upto yield
than did the specimen with two bars. After reaching the yield load, the stiffuess of the
load deflection and moment rotation curves greatly decreased as a plastic hinge
developed near the load point. The model NLRCF captured these phenomena quite
well. Figures 4.16 through 4.19 present the load-deflection and moment- curvature
response of the beams tested. From the figures it is observed that the ultimate load
capacity and ultimate moment from numerical solution agreed well with the test
results. However, the model NLRCF underestimates the deflection significantly and
rotational capacity.
200
160 - 0<>
~ 120
~
- 80 -
40 - ----A-- Test
~ Predicted by NLRCF without strain hardening of steel
- - -e- - - Predicted by NLRCF with strain hardening of steel
o
o 10 20 30 40 50 60 70
Deflection Cmm)
250
200
._." ... G .".-
8'
, 150
3i
~
i:
"
~ 100
.-..-.Test
50
.-..- Predicted by NLRCF without strain hardening of steel
.. -E)_. - Predicted by NLRCF with strain hardening of steel
0
0.01 0.015 0.02 0.025 0.03
0 0.005
Average curvature ( rad)
240
200
160
~
3i
~
.",os 120
0
....l
80
----..-- Test
40 ____ Predicted by NLRCF without strain hardening of steel
.. -e- - - Predicted by NLRCF with strain hardening of steel
0
20 30 40 50
0 10
Deflection (mm)
300
. .0' .
250
200
8
,
~
~
" 150
"a
0
::E 100
--.--Test
50 ---e-- Predictedby NlRCF without strainhardening of steel
- - -e- - - Predicted by NLRCF with strain hardening of steel
0
0 0.0025 0.005 0.0075 om 0.0125 0.015
From Figures 4.16 and 4.18 it is observed that the model NLRCF can predict ultimate
load quite well but underestimates the ultimate deflection. Comparison between load-
deflection curves considering plastic strain hardening and elastic perfectly plastic
idealization presents that the former gives better trend than the latter. From Figures
4.17 and 4.19, it is observed that NLRCF may estimate both moment and curvature
responses with fair degree of accuracy.
•
••••
60
carried out in the modeling of damage and failure phenomena of concrete and RC
structures. To investigate whether the NLRCF is capable of predicting compression
failure of concrete structures, two beams of different sizes and made of two different
concrete types NSC and HSC are studied numerically and compared with the
numerical results obtained by Ozbolt et al. (2000) and also with the test results
obtained from RILEM TC 148SSC. Small beam had a span of 3.60 m, height 200
mm, width 100 mm and large beam had a span of 7.20 m, height 400 mm and width
200 mm. All beams were over-reinforced with the same reinforcement ratio of
approximately 7.3%. To localize compression failure of concrete at the midsection of
the beam the compressive zone was not reinforced over a length of approximately 170
mm in the small beam and 370 mm in the large beam. The small beams are analyzed
for two concrete types and the size effect study is performed only for NSC beams.
Concrete compressive strength used in the analysis was obtained from the uniaxial
test of concrete cylinder. Maximum compressive strain in concrete are considered
0.003 and 0.0045 for NSC and HSC beams respectively. Post peak response ofNSC
beams is much greater than HSC beams. In the analysis the reinforcement is assumed
to be an ideally elastoplastic material. For ribbed steel bars the following properties
are considered: Young's modulus, Es = 222000 MPa, and yield strength, fy = 650
MPa. For wire strands, which were used as a tensile reinforcement in the tests of
small HSC beams, the following properties are considered: Es = 195000 MPa and fy =
1650 MPa the same as used in the experiment. Summary of concrete properties used
in the analysis is presented in Table 4.8.
The results obtained using the model are presented in Table 4.9 along with test results
and numerical results obtained by Ozbolt et al. (2000).
Table 4.8: Summary of concrete properties used in the analysis (Ozbolt et aI., 2000)
Concrete NSC HSC
Ee (MPa) 22000 55000
ft (MPa) 1.89 4.83
f'e (MPa) 22.75 118
61
Table 4.9: Summary of peak load and peak displacement for small beam
Ozbolt et aI., 2000
Test results Predicted Test/Predicted
Concrete Ozbolt et al. NLRCF Ozbolt et al. NLRCF
type Pu 0u Pu 0u Ptest Ptest Otest
kN mm kN mm /Pu lPu lou
NSC 15.5 19.3 15.9 16.2 1.06 1.09 1.03 1.30
HSC 50.2 44.8 63.2 47.2 1.12 1.24 0.89 1.18
Table 4.10: Summary of peak load and peak displacement for NSC beams oftwo
sizes, Small h = 200mm and Lar e h = 400 mm, Ozbolt et aI., 2000
Test results Predicted TestlPredicted
Beam Ozbolt et al. NLRCF Ozbolt et al. NLRCF
Size Pu 0u Pu 0u Ptest Otest Ptest Otest
kN mm kN mm lPu lou lPu lou
Small 21.0 15.5 19.3 15.9 16.2 1.06 1.09 1.03 1.30
Lar e 47.8 56.0 41.1 56.5 50.4 1.14 1.16 1.13 0.95
The numerical results obtained by the model observed failure due to concrete crushing
in the compressive zone of the beam as was obtained in the tests for all over-
reinforced beams. The numerical results obtained using NLRCF are able to predict the
compressive failure ofNSC beams. The numerical and measured results show that the
efficiency of RC beams of different concrete types that fail in compression is not
.
proportional to the uniaxial concrete compressive strength. The predicted response for
HSC beam differs from that of test due to the fact the maximum concrete strain
predicted by the model was 0.00652. The predicted load corresponding to concrete
strain of 0.0045 was 56.2 kN and deflection was 40.3mm which presents reasonably
good agreement with the test results.
To investigate the size effect on the failure load of over-reinforced beam analysis of
two different NSC beam depths (h = 200 and 400 mm) were studied. The average
measured and predicted peak load and corresponding peak displacement are
summarized as presented in Table 4.1O. The values of obtained data are in relatively
good agreement with the average measured values. In all beams failure is caused by
crushing of concrete in the compression zone.
.r~
I' -',,
62
-.-NLRCF
1
0.5
2 150 200 250 300 350 400 45
To present the efficiency of different concrete types the predicted and measured
nominal strength (6Mu/(bh2f~) is plotted as a function of the concrete type where Mu
is the ultimate bending moment. Figure 4.20 presents a good agreement between test
data and numerical results. The test results exhibited nO significant size effect On the
peak load but good agreement is obtained for both the numerical results. This may be
observed from Figure 4.21, which represents the nominal strength as a function of the
beam depth ofNSC.
Theoretical moment-curvature (M-cp) curves for RC sections under flexure with and
without axial load can be derived On the basis of assumptions used in the model. The
curvatures associated with a range of bending moments and axial loads may be
determined using these assumptions and from the requirements of strain compatibility
and equilibrium of forces. The relationship between bending moment, M and
curvature, cpis given by the classical elastic equation as:
"1 •.•
63
EI=M
rp
With increase in moment, cracking of the concrete reduces the flexural rigidity of
the sections. The curvature varies along the length of member because of the
fluctuation of the neutral axis depth. For a given curvature, the position of the neutral
axis depth can be determined by trial and error; i.e. assuming neutral axis depth,
calculating the strain and stress at various points of the sections and equating the
compressive and tensile forces. Once the neutral axis depth is obtained, the moment,
M can be calculated by summing the moments of all forces on the section. The
behavior of the section after cracking is dependent mainly on the steel content.
A doubly RC beam is analyzed using the model to investigate the M-<presponse of the
section. The beam as presented in Park and Paulay (1975) had cross section of
635 x 254 mm. The tension steel was 4- #8 bars and compression steel was 2-#8 bars,
all placed with a 51 mm of cover to the steel centroid. The concrete had cylinder
strength of 20.7 MPa, a modulus of rupture of 2.83 MPa and a modulus of elasticity
of22070 MPa. The steel had yield strength of 276 MPa and a modulus of elasticity of
2 x 105 MPa.
350
300
250
S-
o
3i~ 200
E
~ 150
::E
100
50 -- Analytical
...... , Predicted by NLRCF
0
0 0.01 0.02 0,03 0.04 0.05 0.06
Curvature (radlm)
Fig. 4.22: Moment-curvature relationship ofthe beam section (Park & Paulay, 1975)
1..----___<~...----------------------,
0.8 •
•......•-----'_._..•........... *._.--_.---...•.....~
~
0.2 - ~
o Analytical
...• -- - Predicted by NLRCF
o
o 50 100 150 200 250 300 350
Moment (kN-m)
Fig. 4.23: Variation of un cracked depth ratio of the section at various stages of
loading
II
.
65
0.0045
••• E) ••• Analytical o
-- Predicted by NLRCF
I
1 0.003
.".~
"~
.~ 0.0015
U
~
~
... -.-
." .-
J
.. , ..-
... -'
o
o 50 100 150 200 250 300 350
Moment (kN-m)
Fig. 4.24: Variation of concrete compressive strain of the section at various stages of
loading
From the above figures it is observed that a good correlation exists between analytical
and numerical results. The maximum curvature obtained by the model is less than that
of analytical value due to the difference in extreme fibre compression strain of
concrete. In analysis (Park & Paulay, 1975) maximum concrete compressive strain is
considered 0.004. However in the model the stress-strain curve of concrete is
parabolic for both the ascending and descending branch. The maximum concrete
compressive strain is predicted 0.00254 by the model considering initial tangent
modulus of concrete of22070 MPa.
....;
As
0 0 0
Fig. 4.25: Cross section ofa RC beam (Park & Paulay, 1975)
27.6 -----------------
fT
0.002 0.003 0.004
-----<~.e (mm/mm)
0,0045
~.E~.
_ 0,004
"
• • • • •
.~
o
.~ 0.0035
[
E o Predicted by NLRCF
8
" 0.003
~ • Analytical
8" o
~ 0,0025 o o o o
~
0.002
2 3 4 5 6 7
Beam No.
Fig. 4.27: Variation of maximum concrete compressive strain for the beam sections
0.5
.~ 0.4
,s
~ 0.3
'0
]u 0.2
~ 0.1
o - '--
2 3 4 5 6 7
Be.mNo.
Fig. 4.28: Variation of un cracked depth ratio of the beam sections by NLRCF
68
10
3
8 "". 3
•
4
6
.. --
7 7 6
3 ., .. -
Fig, 4.29: Moment-curvature relationships of beam sections (Park & Paulay, 1975)
69
From the above figures it is observed that there exists a good correlation between the
numerical and analytical results. From Figure 4.29 it is observed that the numerical
ultimate curvature is less than that of analytical one. This may due to the fact that the
analytical moment-curvature is terminated when the extreme concrete fibre strain is
0.004 which is greater than the numerical values obtained from the model as
presented in Figure 4.27. From the theory it is well known that the ductility of single
RC section decreases as the tension steel content is increased, and the presence of
compression steel increases the ductility significantly. This behavior is accurately
predicted by the model and thereby provided the evidence for the efficiency of the
present model.
70
Chapter 5
VERIFICATION OF THE MODEL FOR RC COLUMNS AND FRAMES
5.1 Introduction
An experimental investigation was carried out by Viest et al. (1956) to observe the
effect of sustained loading on the ultimate strength of RC columns. The test results
indicated that the ultimate strength under sustained loading was only 10% below that
for fast loading. A test to failure in a conventional testing machine carried out in about
I hour: the corresponding loading will be referred to herein as "fast loading". Among
45 rectangular columns tested by Viest et al. (1956) 13 tests were made with fast
loading, 12 with slow loading, and 19 with sustained loading. In addition to the type
ofloading concrete strength and eccentricity ofload were major variables.
All columns were of the same size and were reinforced in an identical manner having
a prismatic shaft 20 inch (0.508m) long and 5 x 5 inch (0.127 x 0.127 mm) in cross
section and a 10 inch (0.254 m) capital was provided at each end to accommodate the
loads. The reinforcement was made up of 2 - #4 bars for tension and 2 - # 4 bars for
compression and seven 3/16 inch ties spaced at 4 inch. The concrete cover of the
longitudinal reinforcement was 1/2 inch. Sufficient additional reinforcement was
provided to prevent failure in the capitals. The longitudinal reinforcement was #4
intermediate grade billet steel meeting ASTM A 305-50T for deformations. The
,.
71
average properties of nominal bar were yield strength, 43,300 psi, ultimate strength,
71,500 psi and modulus of elasticity, 27,200,000 psi.
In numerical analysis the modulus of elasticity of concrete has been used 57500...Jfc
where Pc is concrete crushing strength in psi. The cylinder was tested the day the
corresponding column failed. The constitutive model for reinforcement is linear
elastic until yielding and then perfectly plastic. Here only fast loading is considered
for the analysis and the results obtained are presented in Table 5.1 including ultimate
load the column were able to resist and corresponding deformation, the strength of
cylinder, the initial eccentricities measured as the distance from the tension steel to
the load.
Table 5.1: Test and predicted values of ultimate load and deflection at mid height of
column (Vi est et aI., 1956)
As reported by Viest et a!. (1956) that deformation at the ultimate load was obtained
by extrapolation from measurements considered slightly below the ultimate load.
Column 20B3b intended for a sustained load test, failed during the initial load
application. Similarly column 35C4a which was intended for a sustained load test
failed during the initial load application. The arithmetic average of the ratios of
measured to numerical ultimate load for columns subject to fast loading listed in
Table 5.1 is l.OI and the standard deviation is 6.68%. There exists a good correlation
between test and numerical results. The ultimate load obtained by the analysis
considering analyzed cross-section fails when the strains increase to such a values that
either reinforcement yields or concrete crushes and the calculated curvature becomes
very large. The maximum concrete strains obtained by the analysis are l.37-l.48em
which depends on concrete crushing strength and modulus of elasticity of concrete.
Since modulus of elasticity of concrete is not reported by Viest et a!. (1956) Ee used in
the analysis by following ACI 318-95.
The first comparison consists of three columns tested by Abdel Sayed and Gardner
(1975). These three columns were identical except that the concrete cylinder strength
and the eccentricity of the compressive force were varied for each column. Three
columns had a length of3.403m and a square cross-section of 150 x 150mm, and they
were reinforced with 4 - #3 corner bars with yield strength of 447.9 MPa. Table 5.2
presents the strength of columns obtained by both experimentally and numerically.
Table 5.2: Comparison with results by Abdel Sayed and Gardner (1975)
Column t'o e Strength (kN) Ptest /P redicted
elt Lit
No. (MPa) (mm) Test Metwally NLRCF Metwally NLRCF
D- I 30.83 63.5 0.42 22.7 127.0 127.7 127.0 0.995 1.000
D-4 3I.I8 127.0 0.85 22.7 62.4 62.0 63.1 1.006 0.989
D-8 31.04 190.0 1.27 22.7 43.6 40.2 42.2 1.085 1.033
The second comparison consists of columns tested by Drysdale and Huggins (1971).
The two pairs of columns were identical except concrete cylinder strength. All
columns had square cross sections of 127 x 127 mm and reinforced with 4 - #4 corner
bars and hoops of 3.81 mm diameter were spaced 101.6mm centre to centre. All
reinforcement had average yield strength of386.7 MPa, all columns had a slenderness
ratio of 3 I, and axial load having eccentricity equal to 0.2 of the column thickness
was applied to the columns. The strengths of the columns determined both
experimentally and analytically is presented in Table 5.3.
The results in the Tables 5.2 and 5.3 present a very close agreement between the
experimental results and predicted strength by Metwally and by NLRCF.
74
20
18
P = 0.3P.
-lI
I
16
P = O.2Po I
14
12
P = O.IP.
E
~
~
10
"S
2 8
P =0.0
2
P = O.SP.
4-# 3
I-
~
150 mm
c
'I
I~
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Curvature ( rad/m)
Fig. 5.1: Moment- curvature relationship of column section subjected to axial force
.'. 75
It is obvious from Figure 5.1 that the strength of a section improves proportionally
with axial compression upto a certain limit; then the effect reverses and the section
becomes weaker under high compression than without any compression. On the other
hand, the section becomes proportionally less ductile with increasing the axial
compression and the section ductility becomes very limited under large compressive
force. Such behavior is well known and has been demonstrated experimentally. The
behavior of the column D-I (Abdel Sayed and Gardner, 1975) is presented
quantitatively in Figure 5.2. This strength envelope or M-P interaction diagram for
single curvature is based on cross-sectional properties of the column section D-I with
ep =1.0.
1000
Interaction Diagram
800 -
~ 600
.",
15
..-l 400
200 -
o
o 5 10 15 20
Moment (kN-m)
Fig, 5.2: Interaction diagram for the column section D-I
Effect of percentage of steel on the column section is presented in Figure 5.3. From
this figure it is observed that increase in steel content increases the load carrying
capacity of the column section. In Figure 5.3 three steel percentages are considered
with compression to tension steel ratio is equal to 1.00. These steel ratios are 1.26%,
2.5% and 5.0% of the concrete gross section. The effect of these percentages of steel
on the behavior of moment-curvature relationship without axial load and with axial
76
load of P = 0.2P 0, where Po is the axial load carrying capacity of the column section
without flexure, are presented in Figures 5.4 and 5.5, respectively.
1400
Interaction Diagram
1200
1000
~
~ 800
~
."
8
....l 600
400
p ~ 5.0%
200
0
0 5 10 15 20 25 30 35 40
Moment (kN-m)
Fig. 5.3: Interaction diagram of the column D-l at various percentages of steel
(Compression to tension steel ratio = 1.00)
30
p =5.0%
25
6' 20
~ p =2.5%
i:i 15
]
"" 10 p = 1.26%
o
o 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Curvature (rad/m)
Fig. 5.4: Effect of steel on the moment-curvature response of the column D-l
at P = 0.0
, .
77
40
p= 5.0%
35
30
'? 25 p= 2.50%
~
~
20
""
E
15
~ P= 1.26%
10
0
0 0.02 0.04 0.06 0.08 0.1
Curvature (rad/m)
Fig. 5.5: Effect of steel on the moment-curvature response ofthe column D-I
at P = O.2Po
1400
Interaction Diagram
1200
1000
~
~ 800
~
"'0
"
....l'" 600
400
200
f, = 17.24 MPa
0
0 5 10 IS 20 25 30
Moment (kN-m)
Fig. 5.6: Effect of concrete compressive strength on the Interaction diagram ofthe
column D-I for steel ratio, p = 1.26%
•
78
10 ,---------------------------,
8
f, 48.28 MPa
f, 30.83 MPa
E
rr f, = 17.24 MPa
6
~
""
E
0
:E
4 ~ .
0
/ •
25
" .
20
(=48.28 MPa
f, = 30.83 MPa
.f, = 17.24 MPa
5 ~
o
o 0.02 0.04 0.06 0.08 0.1
Curvature (rad/m)
If" "
79
From Figures 5.4 and 5.5 it is observed that an increase in the steel percentage
increases the sectional moment capacity but decreases curvature of the section. This
behavior is also observed at any stage of axial loading, for example, moment-
curvature relationship at 20% of ultimate axial load capacity bear this evidence. The
effect of concrete compressive strength on the interaction diagram of the column
section for steel of 1.26% is presented in Figure 5.6. The effects of concrete strength
on the response of moment-curvature without axial load and with axial load are
presented in Figures 5.7 and 5.8, respectively.
Fifteen two-hinged RC frames were tested by Ernst et al. (1973) to investigate the
performance of frames under vertical and lateral loadings. The test variables were
steel ratios, span of members, grades of steel and steel stress-strain curves. From the
test results it was observed that the connection details of the frame members were
extremely important. Three loading cases were investigated. Case 2 and Case 3 as
reported by the authors, were considered to simulate the response of the frames. In
case 2 a frame was subjected to vertical loading upto failure and in case 3 the lateral
load was increased to fail the frame while carrying a constant vertical load selected as
0.9/1.7 times the ultimate vertical load of a companion frame. The ku (0.9/1.7) factor
essentially provides a vertical working load based on the assumption of a dominant
full live load (i.e., negligible dead load). The A, Band C series were tested in a lOOk
beam testing machine. Case 2 loading was applied to the frames of series A, B, C and
D and case 3 loading was applied on the frame of D series. Cross sectional details and
material properties such as concrete strengths and properties of the steel used in the
frames are summarized in Table 5.4 and Table 5.5, respectively. Figure 5.9 presents
the detail of loads. Table 5.6 presents the test loads by the authors and numerical
results obtained by the model.
80
•
Table 5.4: Cross sectional details of the RC frames (Ernst et aI., 1973)
Nominal dimensions Lon2:itudinal bars
Span Height Beam Column
Frame Beam Column
L h btd btd
(ft) (ft) (in) (in) Outer Inner Outer Inner
A40,A60
C40, C60
209,209H 9
2018,2Dl8H 18
I_ L/3 '1_ L/3 .1. L/3 '1 I_ L/3 .1. L/3 .1. L/3 .,
h h
b
t
III '1
o o
1
0 0
d
t
0 0
o o d
I- .1
Beam section Column section
9' & 12' spans as presented and 18' span = 3 Loads at L/4
Fig. 5.9: Frame geometry and load details (Ernst et aI., 1973)
81
Table 5.6: Predicted and test loads and types offailure (Ernst et a!., 1973)
Vertical Load
Test loads (kN) Predicted (kN)
Load Yield attained at Yield attained at
Frame Maximum Maximum
case W. Mid E. W. E.
Load Load Midspan
knee span knee knee knee
A40 2 70.9 67.1 51.5 62.7 67.8
76.4 67.2 67.2
A60 2 88.8 77.2 77.2 88.8 82.8
100.0 82.2 82.2
B40 2 82.5 41.6 75.7 54.0 76.4 50,8 75.6 50.8
B60 2 93.0 64.2 86.4 64.2 80.8 51.8 77.6 51.8
C40 2 72.9 62.1 40.6 62.1 72.6 70.4 55.4 70.4
C60 2 77.2 77.2 61.6 74.0 96.6 86.1 64.4 86.1
2D9 2 94.2 87.0 62.6 87.9 94.3 83.2 51.6 83.2
2DI2 2 66.2 66.2 57.1 66.1 68.0 61.3 42.6 61.3
2Dl8 2 46.4 44.7 34.9 44.2 46.8 42.8 30.2 42.8
•
82
Lateral Load
Test loads (kN) Predicted (kN)
Yield attained Vertical
Load Yield attained at
Frame Maximum Maximum at load
case
load W. Mid load W. Mid (kN)
knee span knee span
2D9H 3 13.2 11.5 10.9 12.7 11.5 --- 49.8
2Dl2H 3 13.7 13.2 13.1 12.8 12.7 --- 35.1
2Dl8H 3 14.1 11.9 11.1 11.9 11.3 --- 24.6
From Table 5.6 it is observed that there is a good correlation between test and
numerical results for both the load case 2 and load case 3. For the frames A60 and
C60, ultimate loads obtained are somewhat larger than the corresponding test results.
The vertical load vs. midspan deflection responses of the frames is presented in
Figures 5.10 to 5.13 and lateral load vs. lateral displacement of the frames is
presented in Figure 5.14 in comparison with the test results.
120
100 A60
80
--- ---
--------
~
-g
60 --- --- --- A40
.2
~
40
- Predicted by NLRCF
20 ---- Test
0
0 10 20 30 40 50 60
Midspan deflection (mm)
Fig. 5.10: Load vs. midspan deflection of the RC frame (Ernst et aI., 1973)
83
100
,,
B60
80
----
60
~
-g
""
"Ii
~ 40
20
- Predicted by NLRCF
---- Test
o
o 10 20 30 40 so
Midspan deflection (mm)
Fig. 5.11: Load vs. midspan deflection of the RC frame (Ernst et aI., 1973)
120
100 C 60
80
--- ---
~
-g
--
60
""
"Ii -' C 40
~
40
- Predicted by NLRCF
20
---- Test
0
0 10 20 30 40 so
Midspan deflection (mm)
Fig. 5.12: Load vs. midspan deflection of the RC frame (Ernst et aI., 1973)
84
100
--
80
~ 60
~
----- ------- ------
~ 40 --- -----
20
- Predicted by NLRCF
---- Test
o
o 20 . 40 60 80 100 120
Midspan deflection (mm)
Fig. 5.13: Load vs. midspan deflection of the RC frame (Ernst et aI., 1973)
15
3
Predicted by NLRCF
Test
o
o 10 20 30 40 50 60 70 80 90
Lateral deflection (mm)
Fig. 5.14: Lateral load vs. lateral deflection of the RC frame (Ernst et aI., 1973)
85
From Table 5.6 it is observed that for the frames A60 and C60, ultimate loads
obtained are somewhat larger than the corresponding test results. This may due to the
fact that in case of A60 bar fractures was observed at W. knee while for frame C60
bar fracture was observed at E. knee of the frame. For the load case 3, it is observed
that the model simulates the ultimate load and W. knee yield but midspan yielding is
not observed. From the test results it is observed that midspan yielding was observed
after the ultimate load.
From the above figures it is observed that the numerical results obtained using the
model NLRCF presents a good agreement with the test results. Deflection obtained
from the test results is greater than that of the numerical results and this pattern varies
widely at load close to the ultimate load.
The influence of masonry infill panels on the seismic performance of RC frames was
studied by an experimental investigation conducted by Mehrabi et al. on 12 one-half
scale frame specimens in 1996. Out of 12 frames, one frame was bare frame which is
subjected to monotonic lateral load upto failure. This specimen is considered here for
analysis using the model NLRCF. The specimen I, reported by Mehrabi et al. (1996),
is a single story single bay frame having a span of 2.337m and height of 1.5365 m.
8 -# 4 deformed bars were used as main reinforcement for column sections with #2
plain bars as a tie with spacing 12.7 mm centre to centre. In the beam section, 4 - # 5
deformed bars were used as flexural reinforcements and # 2 plain bars as a stirrup
with a spacing of 76 mm centre to centre. Load arrangements; beam and column
sections are presented in Figure 5.15. Table 5.7 summarizes the material properties
used in the test of specimen I and used in the model.
86 'T
147 kN 147 kN
1,152 mm 178mm
Load .1 I- III
E
E
~
N
N
l L-.
Beam
-#2
/,4-#5
II~
8 -# 4 Column
#2
Fig. 5.15: Cross-sectional details and loading procedures for the specimen 1
(Mehrabi et a!., 1996)
B. Reinforcing bar
Nominal dia, Yield strength, MPa Ultimate strength, MPa
Bar size
mm(inch) (ksi) (ksi)
#2 6.35 (0.25) 367.6 (53.3) 449.6 (65.2)
#4 12.7 (0.50) 420.7 (61.0) 662.1 (96.0)
#5 15.9 (0.625) 413.8 (60.0) 662.1 (96.0)
The lateral load was applied on the frame as presented in Figure 5.15 with a total
vertical load of 294 kN on to the top of the column first. From the test results it is
observed that the bare frame exhibit a fairly ductile behavior and first hinge is
developed in the longitudinal reinforcement in the bottom sections of the column and
later in the flexural reinforcement at the end section of the beam. Concrete crushing
was observed at the upper end of the windward column and lower end of leeward
column at the maximum load of 107 kN and a lateral displacement of about 48.3 mm,
after which the lateral resistance remained more or less constant.
••
..-;g~
. \
87 ,
From the analysis using the model it is observed that first flexural crack developed at
the bottom section of the column of windward side at a load of 27 kN and a lateral
displacement of 2.2 mm. Yielding of the frame is first detected in the longitudinal
reinforcement on the bottom section of the column at a lateral load of 70.2 kN and a
displacement of 11.9 mm. Concrete crushing is observed first at the lower end of the
leeward column then at the lower end of windward column and the maximum lateral
load obtained by the model is 80.3 kN with a displacement of 15.7 mm.
From the analysis it is also observed that the model fairly simulates the ductile
behavior of the frame as in the test but load and deflection obtained by the model are
le'sser than that of the test results, This might be due to the fact that strain hardening of
steel is not considered. If the strain hardening increases it increases the load carrying
of the frame. The crushing of concrete and steel yielding of the bottom of the column
sections are observed by the analysis before the yielding of the longitudinal
reinforcement of the ends of the beam section when elasto-plastic response of steel is
considered.
•
...~
88
B. Reinforcing bar
Yield strength (MPa) 338.5
Modulus of elasticity (MPa) 200000
203 mm
I- 165 mm 'I
Load I- 'I
6 gage
6-#34_#3
~I!I 6 gaee
Column section
Beam section
Fig. 5.16: Cross-sectional details and loading procedures for the model I
(AI-Chaar et aI., 2002)
The lateral load was applied on the frame as presented in Figure 5.16 with a single
actuator having a capacity of 222 kN for the model I. The specimen was laterally
braced to eliminate out-of-plane movements. From the test results it is observed that
the experimental peak load was 28.9 kN with a lateral displacement of 12.4 mm. The
displacement at half peak load was 4.22 mm. In the RC bare frame, cracks began to
form in the beam-column joint on the tension side at approximately 1110 of the total
displacement. This was due to the discontinuity of the bottom longitudinal
reinforcement in the beam section of the frame. Next a small tension crack appeared
at the top of the compression column, shortly followed by both a tension crack near
the base of the tension column and a shear crack near the base of the compression
column.
89
From the analysis using the model it is observed that the strain hardening effect of
steel play a significant role for the development of collapse mechanism. The first
hinge developed at the tension side of the compression column at a lateral load of 18.4
kN with a lateral displacement of 5.8 mm. Then second hinge is formed at the tension
face of the base ofthe tension column at a load of 19.6 with a displacement of6.3 mm
after that numerical convergence is not occurred if the elasto-plastic behavior of steel
is considered. Displacement at load of 14.45 kN which is the half peak load of the
experiment is 4.5 mm which is very close to the experimental results. Using 30%
strain hardening of steel, a third hinge is developed on the beam column joint at the
leeward side and onset of the fourth hinge on the windward side of the frame for a
load of 24.6 kN with a displacement of 8.0 mm. Figure 5.17 presents the sequence of
yielding of the steel reinforcement.
4 3
Chapter 6
CONCLUSIONS AND RECOMMENDATIONS
6.1 General
6.2 Conclusions
The objective of this study was to develop reliable and computationally efficient finite
element model for the analysis of RC beams, columns and frames under monotonic
static loading conditions. Comparisons between the results obtained using the model
NLRCF (Nonlinear numerical model for analyzing Reinforced Concrete Frame)
developed and the experimental data present very good correlation in the event of
load-deflection and moment-curvature response of sections. From the analyses using
the model NLRCF following conclusions may be summarized:
.
I
'
•I .
91
o The numerical model developed in the study can successfully simulate load-
deflection and moment-curvature responses ofRC beams, columns and frames
under loads upto collapse.
o The load carrying capacity ofRC column increases but the ductility is reduced
with the increase of steel content. However, the ductility increases as the
concrete compressive strength is increased.
The beam element, used in developing the nonlinear numerical model considers
stress-strain relationships of the constituent materials like steel and concrete. The
cracked and uncracked stiffnesses of concrete and yielding of reinforcing steel in a
RC member has been taken into account and is implemented in the computer program
NLRCF. This model (NLRCF) has several opportunities for future enhancements.
• The model NLRCF has been developed for rectangular RC sections. Software
for other shapes of RC sections may be included in the program.
92
• The beam element that has been implemented in the model does not account
for deformations due to shear and torsion. So further development may be
carried out to include these effects.
I •.
•
93
REFERENCES
Abdel Sayed, S. I., and Gardner, N.J., (1975),"Design of Symmetric Square Slender
Reinforced Concrete columns under Biaxially Eccentric Loads", Reinforced Concrete
columns, SP-50, American Concrete Institute, Detroit, pp 149-164.
ACI Committee 318, "Building Code Requirements for Reinforced Concrete (ACI
318-71 )," American Concrete Institute, Detroit, 1971, 78 pp
ACI Committee 318, "Building Code Requirements for Reinforced Concrete (ACI
318-83)," American Concrete Institute, Detroit, 1983, Illpp
Bashur, F.K, and Darwin, D. (1978), "Nonlinear Model for Reinforced Concrete
Slabs" Journal of Structural Division, ASCE, Vol. 104, No. ST!, pp 157-170.
Bresler, B. and Scordelis, A.C. (1963), "Shear Strength of Reinforced Concrete
Beams", Journal of ACl, Vol. 60, No. I, pp 51-72.
DIANA, version 6.1(1996), Dept. of Engg. Mech. and Information Technol., TNO
Building and Construction Research, Delft, The Netherlands.
94
Drysdale, RG., and Huggins, M.W. (1971), "Sustained Biaxial Load on Slender
Concrete Columns", Journal o/the Structural Division, ASCE, Vol. 97, ST5, pp 1423-
1443.
Ernst, G.C., Smith, G.M., Riveland, A.R and Pierce, D.N. (1973), "Basic
Reinforced Concrete Frame Performance Under Vertical and Lateral Loads", ACI
Journal, Vol. 70, No.4, pp 261-269.
Hognestad, E. (1951), "A Study of Combined Bending and Axial Load in Reinforced
Concrete Members", University of Illinois Engineering Experiment Station, Bulletin
Series No. 399, Bulletin No.1.
Lin, C.S. and Scordelis, A.C. (1975), "Nonlinear Analysis of RC Shells of General
Form", Journal o/Structural Division, ASCE, Vol. 101, No. ST3, pp. 523-538.
Lin, T.Y. and Burns, N.H. (1982), Design 0/ Prestressed Concrete Structures, 3'd
edition, John Willey and Sons, New York.
Mansur, M.A., Chin, M.S., and Wee, T.H. (1997), "Flexural Behavior of High-
Strength Concrete Beams", ACI Structural Journal, Vol. 94, No.6, pp 663-674.
Mehrabi, A.B. and Shing, P.B. (1997), "Finite Element Modeling of Masonry -
Infilled RC Frames", Journal o/the Structural Division, ASCE, Vol. 123, pp 604-613.
Mehrabi, A.B., Shing, P.B., Schuller, M.P. and Noland, J.L. (1996)," Experimental
Evaluation of Masonry-Infilled RC Frames", Journal 0/ Structural Engineering,
ASCE, Vol. 122, No.3, pp 228-237.
Nilson, A.H. (1972), "Internal Measurement of Bond Slip", Journal of ACI, Vol. 69,
Title No.7, pp 439-441.
Ozbolt, J., Mestrovic, D., Li, Y.-J. and Eligehausen, R. (2000), "Compression
Failure of Beams Made of Different Concrete Types and Sizes", Journal of Structural
Engineering, ASCE, Vol. 126, NO.2, pp 200-209.
Pfrang, E.O., and Siess, c.P. (1964), "Predicting Structural Behavior Analytically",
Journal of the Structural Division, ASCE, Vol. 90, pp 99-111.
Sawko, F. and Rouf, M.A. (1985), "A proposed Numerical Model for Structural
Masonry", Masonry International, NO.5.
Selna, L.G. (1969), "Creep, Cracking and Shrinkage in Concrete Frame Structures",
Journal of the Structural Division, ASCE, Vol. 95, No. ST12, pp. 2743-2761.
Stevens, N.J., Uzumeri, S.M., Collins, M.P. and Will, G.T. (1991), "Constitutive
Model for Reinforced Concrete Finite Element Analysis", ACI Structural Journal,
Vol. 88, No. I, pp 49-59.
Viest, I.M., Eistner, R.c. and Hognestad, E. (1956), "Sustained Load Strength of
Eccentrically Loaded Short Reinforced Concrete Columns", Journal of ACI, Vol. 27,
No.7, pp 727-756.
97
APPENDIX-I
INTEGERP, G, F, H, FAIL
COMMON 1 BLK61 ESS(I 00, I O),ESP(I 00, I O),EYY(I 00, I O),EU(I 00, I 0)
*,QLL(300)
COMMON 1 BLK8 1 IQN(I OO),QSL(I 00, I O),QSR(I 00, I O),QD( I 00, I 0)
*,QAS(lOO,IO)
COMMON 1 BLK9 1 ESTNI(300,10), ESTN2(300,10),ES(300,10)
COMMON/BLKIOI AS(300, I 0),IQNN(300),QDX(300, 10)
COMMON/BLKIIITSC
COMMON/BLKI2/ESS I ,ESP I ,EYYI ,EUl
COMMON 1 BLKI3/ESSN(300, I 0),ESPN(300, I 0),EYYN(300, I 0),EUN(300, I 0)
98
C INITIALISE VARIABLES
IS23=9
ISREE = ISRE
TOTRES~O.O
REDIS = 0.0
NCHING~O
IXX~O
INAL~O
INALL~O
NOUT~l
NSTOP~O
Bl ~ 9999.0
RRI ~ 0.0
TSC~FTCIYM
MM=3*M+3
NDL=NDL-l
MKQ~M/2+4
IF(ISYMM .EQ. 0 .OR. M .LE. 16) MKQ = M
NODE~M+ I
99 j
EEl = 2.0*FYNM
ISTIFF ~ I
DO 1l0J~ I,MM
FO(J) = 0.0
FCJ(J) ~ 0.0
FFC(J) ~ 0.0
FT(J) ~ 0.0
FLL(J) ~ 0.0
D(J) ~ 0.0
DC(J) ~ 0.0
DT(J) = 0.0
DLL(J)~ 0.0
DCONV(J) ~ 0.0
FCONV(J) = 0.0
110 CONTINUE
NC~O
AXIS(NODE) ~ 1.0
DO 130 I~I,M
PRE(I) = 0.0
SS(I) ~ 0.0
SSTR(l) ~ 0.0
SHST(l)= 0.0
SMF(l) ~ 0.0
1ST (I) ~ 0
AXIS(l) = 1.0
STRAIN(l) = 0.0
CURV(I) ~ 0.0
U(I) = 0.0
THETA(I) = 0.0
YC(I) ~ 0.0
CCQNI(I) ~ 0.0
CCQN2(I) = 0.0
DNI(I) = 1.0
DN2(I) = 1.0
DO 120J~ 1,3
ZO(I,J) = 0.0
ZI(I,J) = 0.0
120 CONTINUE
130 CONTINUE
NMll=O
DO 5566 INI~I,NBE
TXC I ~TX(IN I)
BXCI=BRX(INI)
NMI~NELX(INI)
DO 5567 IN2=I,NMI
AREA(IN2+NMI1)~TXCI*BXCI
EI(IN2+NMII)=BXCI*TXCI **3112.0*YM
EY(IN2+NMI1)~YM
T(IN2+NMII)~TXCI
BR(IN2+NMII)~BXCI
5567 CONTINUE
NMII=NMI I+NMI
5566 CONTINUE
AREA(NODE) = AREA(IN2-1)
AXIS(NODE) = 1.0
T(NODE) ~ T(IN2-1)
100 t
BR(NODE)~BR(IN2-1 )
C
C EVALUATE GEOMETRY
C
IF(NGEOM .EQ. 4) GO TO 91
IF (NGEOM .EQ. 5) GO TO 93
CALL GEOM (A,B,M,NGEOM,BI)
GO TO 92
93 X(I) = 0.0
Y(I)~O.O
K=O
DO 94 I = I,NBE
IF(I .GT. I) K = K+NELX(I - I)
KQI=NELX(I)
YY I ~BX(I)IKQ 1
XXI =AX(I)IKQ I
XYLI ~DSQRT(XXI**2+YYI**2)
QXI=XYLI
QXLl~QXl/2.0
DO 501 II=I,KQI
IQ=IQN(I)
IQNN(II+K)~IQ
QLL(II+K)=QL(I)
DO 502 IK=l,IQ
QD X(II +K,IK)~Q D(I,IK)
ESSN(II+K,IK) = ESS(I,IK)
ESPN (II +K,IK )=ESP(I,IK)
EYYN(lI +K,IK)~ EYY (I,IK)
EUN (II +K,IK)=EU(I,IK)
ES(II+K,IK)~O.O
ESTN1 (II+K,IK)=O.O
ESTN2(1I+K,IK)~0.0
AS(lI+K,IK)~O.O
IF(QXLI .GE.QSL(I,IK) .AND. QXLI .LE. QSR(I,IK)) THEN
AS(II+K,IK)=QAS(I,IK)
ENDIF
WRITE( 6,66)(II+K),IK,AS(II+K,IK), QDX(lI +K,IK)
66 FORMAT(30X,'AS('15,','15,') = ',F14.10,IOX, 'DEPTH = ',FIO.6)
502 CONTINUE
C T(NODE) ~ T(M)
C
C
BR(NODE)~BR(M)
DO 77003 I ~ I,M
•
C AREA(I) ~ T(I)*BR(I)
C EY(I) ~ BR(I)*T(I)**3/l2.0
C EI(I) ~ EY(I)*YM
C77003 CONTINUE
101 '. ,.
GO TO 92
91 CALL COORD (A,B,M,NGEOM,RRI)
92 WRITE (6,8010)
WRITE (6,8020) (I,X(I), Y(I), U(I), THET A(I),GA(I),XL(I), T(I),BR(I),
*1,1~ I,M) .
8010 FORMAT(/IOX,'GEOMETRIC PROPERTIES OF STRUCTURE'/
1 IOX,30('*')// IX, 'NODE',7X,'X'
1,ISX,'Y', ISX,'U', 12X,'THET A', 12X,'GA', ISX,'XL'
2,8X,'THICK',SX,'WIDTH',SX'EL NO')
8020 FORMAT (2X,IS,8FIS.S,IS)
IF (NDEAD.EQ.O) GO TO 170
C
C EVALUATE DEAD LOAD
C ADD STRUCTURE OWN WEIGHT
INALL = INALL + I
IF(ISRE .EQ. 2) ISREE = ISREE-2
TXIFL ~ 0.0
TXA=O.O
IF (IDLOAD .EQ. I .OR. IDLOAD .EQ. 4)GO TO ISS
IF(IDLOAD .EQ. 3) GO TO IS7
C
C IDLOAD ~ I - -> SELF + INFTLL LOAD APPLIED UNIFORMLY OVER ARCH
C " ~ 2- -> FOR SELF WT. OF ANY STRUCTURE
C " =3- -> APPLYING D.L. AS NODAL LOADS.
C " ~4- -> APPLYING INFILL LOAD AS CONC, / DIST, LOAD ON ARCH
C
C DEN = DEN*BR(I)
DO lSI 1= I,M
FOOX ~ DEN*BR(I)*XL(I)*T(I)/2.0
TXA ~ TXA + 2.0*FOOX
TXXI = FOOX*XL(I)*DCOS(GA(I))/6.0
II ~ 3*(1-1)
FO(lI+ 2)~FO(II+ 2)+FOOX
FO(II +S)~FO(II+S)+ FOOX
FO(lI+ 3)~FO(II+ 3)+ TXXI
FO(II+6)~FO(II+6)- TXX 1
lSI CONTINUE
GO TO IS6
C
C IDLOAD ~ I - -> DEAD WTS. APPLIED UNIFORMLY
140 CONTINUE
TOTX~O.O
DO 141 1~I,MM,3
FOOX~FO(l)
FO(l)~- FO(I+ I)
FO(I+l)~-FOOX
141 CONTINUE
IF(IDLOAD .EQ. 4) GO TO 159
GO TO 156
DO 161 I~I,MM,3
FOOX~FO(l)
FO(I)~FO(I+ I)
FO(I+ I)~FOOX
161 FO(I+2)=-FO(I+2)
DO 162 I~I,MM
FO(I)=-FO(l)- F I(I)
162 Fl(I)~O.O
DO 143 I=I,NJR
MMN ~ 3'(NJ(I)-I)
IF(NX(I) .EQ. 1) FO(MMN+I)~O.O
IF(NY(I) .EQ. I) FO(MMN+2)~0.0
IF(NM(I) .EQ. I) FO(MMN+3)~0.0
143 CONTINUE
142 FOOX=FLOAT(NDL+I)
DO ISOI=I,MM
IF(IDLOAD.NE.3) FO(I) = -FO(l)
FO(l) ~ FO(l)/FOOX
TOTRES = FO(l)'FO(l) + TOTRES
F1(I) = FO(I)
•.
. '
103
150 CONTINUE
TOTX~O.O
DO 77001 1~2,MM,3
77001 TOTX ~ TOTX + FO(I)
TOTRES ~ DSQRT(TOTRES)
WRITE (6,99985) TOTX, TXIFL,TXA
WRITE (6,8030) (J,FO(J*3-2), FO(J*3-1),FO(J*3),J,J=I,NODE)
GOTO 300
8040 FORMAT (13/(4X,6FI2.0))
8050 FORMAT(4X,6FI2.0)
8030 FORMAT(lIX,I5,3X,3FI2.5,I5)
8261 FORMAT(lIX,I5,3X,3FI2.5)
C
C APPLY LIVE LOAD
C
170 DO 180 I=I,MM
FO(l) ~ 0.0
180 CONTINUE
IF(NLM .EQ. 0) GO TO 182
M~O
DO 555 IJK=I,NBE
A~AX(IJK)
B~BX(IJK)
M~NELX(IJK)
BR=BRX(IJK)
IF(IJK .GT. I) M~M+NELX(lJK-I)
CALL LOOF (A, B, M, MM, AA, LC, LL, LT!, LT2, BR, BI,
* NGEOM, ROAD, T, LOmS, 2)
555 CONTINUE
TOTRES~O.O
DO 181 I=I,MM,3
FOOX~FO(l)
FO(l)~ FO(I+ I)
FO(I+ I)~ FOOX
FO(I+2) = - FO(I+2)
181 CONTINUE
182 IF(NJL .EQ. 0) GO TO 185
C
C ADD FORCES APPLIED AT NODES
C
DO 184 I~I,NJL
MMN ~ 3*(NJJ(I)-I)
FO(MMN+ I) = FO(MMN+ 1)+FX(I)
FO(MMN+2) ~ FO(MMN+2)+FYY(I)
FO(MMN+3) = FO(MMN+3)+FM(l)
184 CONTINUE
185 CONTINUE
C
C MAKE FORCES ZERO ALONG RESTRAINED D. O. F
DO 183 I~I,NJR
MMN = 3*(NJ(I)-I)
IF(NX(I) .EQ. 1) FO(MMN+1) ~ 0.0
IF(NY(I) .EQ. 1) FO(MMN+2) = 0.0
IF(NM(I) .EQ. I) FO(MMN+3) = 0.0
183 CONTINUE
TOTRES~O.O
•
104 ..... ~
DO 190 I~I,MM
TOTRES ~ FO(I)*FO(1) + TOTRES
FI(I) ~ FO(I)
190 CONTINUE
TOTX~O.O
TOTH=O.O
TOTM=O.O
DO 191 I~I,MM,3
TOTX ~ TOTX+FO(I)
TOTH ~ TOTH +FO(I+ 1)
191 TOTM = TOTM + FO(I+2)
TOTRES ~ DSQRT(TOTRES)
WRITE (6,99983) TOTRES,TOTH
WRITE (6,99980)
WRITE (6,8261) (J,FO(3*J-2), FO(3*J-I), FO(3*J),J=I,NODE)
201 INAL = INAL + I
INALL ~ INALL + I
ISTIFF = 0
210 ISTIFF ~ ISTIFF + I
NC~O
IF(ISRE .EQ. 2) ISREE = ISREE-2
IF (NCONT .EQ. 2 .OR. NCONT .EQ. 4) GO TO 300
IF (NDEAD .EQ. 0 .AND. INAL .EQ. I) GO TO 300
C
C EVALUATE NEW GEOMETRY
C
00211 I~I,M
211 CALL NEWGEOM(I,NGEOM)
C
C FORM ELEMENT STIFFNESS MATRIX AND ADD TO TOTAL STIFFNESS MATRIX
C
300 IF (NCONT .EQ. 4 .AND. INALL .GT. I) GO TO 410
IF (NCONT .EQ. 2 .AND. ISREE .LT. 0) GO TO 410
DO 320 K~I, MM
DO 320 J~I,6
320 S(K,J) = 0.0
DO 350 I~I,M
IF (NCONT .EQ. 3) GO TO 352
IF (NCONT .EQ. 1 .AND. ISRE .EQ. 2) GO TO 352
GOTO 353
352 AXIS(I) = 1.0
STRAIN(1) ~ 0.0
CURV(I) ~ 0.0
353 INN=I
CALL STIFF (AXIS(1),STRAIN(1),EE 1,CURV(I),T(I),XL(I),YM,EI(I),
I AREA(1), GA(I), SII, S12, S21,S22,I,MKQ, IPALL, INAL, NDEAD,
2 CCQNI(1),CCQN2(I),ISI3,DNI(1),DN2(1), IS23,NC,INN)
L = 3*1
351 D0340J=I,3
F ~ L+J-3
H ~ L+J
DO 330 K=I,3
P ~ K+3-J+1
G=K-J+l
8056 FORMAT (2X,3(5I5,5X))
S(F,P) ~ S(F,P) + SI2(J,K)
105
IF (G.LE.O) GO TO 330
S(H,G) ~ S(H,G) + S22(J,K)
S(F,G) = S(F,G) + SII(J,K)
C WRITE(6, *) 'MAIN', I,P,G,F,H,S(F,P),S(H,G),S(F,G)
330 CONTINUE
340 CONTINUE
350 CONTINUE
8070 FORMAT (/40X,'TOTAL STF MTX'/(4X,6FI2.0))
8071 FORMAT (/40X, 'EL.STF.MTX.'/(I0X,6FI4.2))
8072 FORMAT (I0X,6FI4.2)
C
C APPLY BOUNDARY CONDITIONS
C
DO 370 I=I,NJR
MMN = 3*(NJ(I)-I)
IF(NX(I) .EQ. 0) GO TO 377
IKK~6
DO 371 IK=2,6
37 I S(MMN+ I ,IK)~O.O
IF(MMN+ I .LT. 6) IKK ~ MMN+ I
DO 372 IK = I,IKK
S(MMN+2-IK,IK)~0.0
372 CONTINUE
377 IF(NY(I) .EQ. O)GO TO 378
IKK~6
DO 373 IK=2,6
373 S(MMN+2,IK)=0.0
IF(MMN+2 .LT. 6) IKK=MMN+2
DO 374 IK=I,IKK
S(MMN+3-IK,IK)~0.0
374 CONTINUE
378 IF(NM(I) .EQ. 0) GO TO 370
IKK=6
DO 375 IK ~ 2,6
375 S(MMN+3,IK)=0.0
IF (MMN+3 .LT. 6) IKK=MMN+3
DO 376 IK=I,IKK
S(MMN+4-IK,IK)~0.0
376 CONTINUE
370 CONTINUE
C
C DECOMPOSE AND SOLVE EQUATIONS
C •
404 CALL DECBND(MM, 6, FAIL)
IF (FAIL .EQ. I) GO TO 950
cc WRITE (6,88888)
410 DO 405 I=I,MM
405 D(I)=O.O
CALL SL VBND(MM,6)
8100 FORMA T(/40X,'STF EQN'/(I4,6F 13.2,3X,Di1.3,FI 0.3))
NC=NC+ 1
C
C MODIFY NODAL COORDINATES
C
IF(NCONT .EQ. 2 .OR. NCONT .EQ. 4) GO TO 444
DO 411 I~I,NODE
X(I) ~ X(I)+D(3*1-2)
Q.'
106
C
420 DO 430 I~I,M
INN=I
CALL VOL(XL(I),GA(I),I,CURV(I), STRAIN(I),AXIS(I),T(I),
I EE I, ZO(I, I ),ZO(I,2),ZO(I,3),ZI(I, I), ZI(I,2),ZI(I,3), FY,BR(I)
* ,PP,EMAX(I),RA TIO(I),NC,MKQ,IPALL,NCONT ,DNI(I),DN2(I)
* ,CCQN 1(I),CCQN2(I),EMXN I (I),EMXN2(I),A3(I),A4(I),A6(I),INAL,
*INN)
DO 88810 J=I,3
PPL(I,J) ~ PP(J)
PPR(I,J) ~ PP(J+3)
c WRITE(8, *) 'MAIN',I,J,PPL(I,J),PPR(I,J)
88810 CONTINUE
430 CONTINUE
8110 FORMAT(l60X,'ZO,ZI BY VOL'/(1X,6Fl1.6,2D13.6,2FI3.8))
C
C EVALUATE TOTAL NODAL FORCES
C
DO 421 J=2,M
DO 421 K=I,3
I~J*3-3+K
421 FT(I) ~ ZO(J,K)+ZI(J-I,K)
DO 431 J=I,3
FT(J) ~ lO( 1,J)
431 FT(MM+J-3) = ZI(M,J)
C
C MAKE FORCES ZERO ALONG REST, D,O,F,
C
DO 380 I=I,NJR
MMN ~ 3*(NJ(I)-I)
IF(NX(I) .EQ. 1) FT(MMN+l) = 0.0
IF(NY(I) .EQ. I) FT(MMN+2) ~ 0.0
IF(NM(I) .EQ. I) FT(MMN+3) ~ 0.0
107
CCQNI(I) ~ CCQNI(I)*T(I)IEEI
C CCQNI(I) = CCQN1 (I)*AXIS(I)
CCQN2(I)~CCQN2(I)*T(I)IEE I
C CCQN2(I)=CCQN2(I)* AXIS(I)
FI(I)= FIRRC*FI(I)
IF(I .EQ. 1) TOTRES = TOTRES * DABS(FIRRC)
963 FO(I) ~ FI(I)
DqI) = 0.0
439 CONTINUE
cc426 WRITE (6,88896)
426 CONTINUE
DO 432 I~I,MM,3
IF(I.GT. 3*(MKQ+I)) GO TO 432
DTU ~ DT(I)* I 000.0
DTV = DT(I+ 1)* 1000.0
DTT ~ DT(I+2)* 1000.0
DLU ~ DLL(I)*1000.0
DLV = DLL(I+ 1)*1000.0
DLT ~ DLL(I+2)* I 000.0
IPD ~ 1/3+1
IF(lPD .LT. 31 .OR.IPD .GT. 31)GO TO 55009
c IF(IPD .LT. 31 .OR.IPD .GT. 31)GO TO 432
432 CONTINUE
CC428 WRITE(6,88890)
428 CONTINUE
WRITE( 6,88893) (I,(PPL(I,J),J~ 1,3),PPR(I,3 ),ECCL(I),ECCR(I)
*,CCQN 1(I), CUR V I (I), CCQN2(1),EMXN 1(I),EMAX I (I),EMXN2(1),
* DNI (I),AXIS(I),DN2(I),STR I (I),RA TIO(I),I=I ,M,45)
c WRITE (6,33333)
WRITE( 6,33334 )(I,(ES(I,J),ES TN 1(I,J),ESTN2(I,J),J~ I ,IQNN(INN)),
*I=I,M,45)
DO 77010 I~I,M
SS(I) ~ DABS(PPL(I,2))/AREA(I)
SSWI=O.O
DO 55005 N=I,IQNN(INN)
SSW I ~AS(I,N)+SSWI
55005 CONTINUE
PW~SS W I I( AREA(I) *0.82)
109
FACT~ DABS(PPL(I,2)*T(I)IPPL(I,3))
IF( FACT .GT. 1.00) FACT~1.0
SS 1~6.89655 *( 1.9*SQRT(FY/6.89655)+2500*F ACT*PW)
C SSI ~1 000*(0.14*SQRT(FYIl000)+ 17.2*SSWI*DABS(PPL(I,2)IPPL(I,3))
C */BR(I))
C SSTR(I)=0.14*1000*SQRT(FYIl000)+ 17.2*SSWI*PPL(I,2)/
C *(PPL(I,3)-PPL(I, 1)*3 .0*T(I)/8.0)/BR(I)*1 000
SSTR(I)= I 000*0.17*( 1+0.073 *DABS(PPL(I, 1))/(BR(I)*T(I)))
IF(SS(I) .LT. SS I) GO TO 77010
SHST(I)~SAS*ESS I*EYY1/SP/BR(I)
IF(SAS .EQ. 0.0 .OR. SP .EQ. 0.0) SHST(I) ~O.O
SHl ~SHST(I)+SS 1
C IF(SS(I) .LE. SSTR(I) ) GO TO 770 II
C IF(SS(I) .LE. SHI) GO TO 77012
IF(SS(I) .GT. SH1) GO TO 77013
IF(SS(I) .LE. SSTR(I)) GO TO 77010
IF(SS(I) .LE. SHI) GO TO 77010
IST(I) = 4
SMF(I) ~ SSTR(I)/SHI
WRITE(6,77099)SMF(I)
GO TO 77020
770ll IST(I) ~ I
GO TO 77020
77012 IST(I) = -22
GO TO 77020
77013 1ST (I) = 333
SMF(I) ~ SS(I)/SH I
C77020 WRITE (6,77097)I,PRE(I),SSTR(I), SS(I), SMF(I),IST(I),I
77020 WRITE (6,77097)I,SHST(I),SSTR(I), SS(I), SMF(I),IST(I),I
IF (SMF(I) .GT. 1.0) GO TO 66116
77010 CONTINUE
77009 CONTINUE
C427 WRITE(6,88894)
427 CONTINUE
IF(NCONT .GT. 2) GO TO 962
IF(IDCONV .EQ. I) GO TO 425
IF(CONV .GT. CONVLT) GO TO 425
962 IF(INAL. EQ. LODINC) GO TO 960
IF(INAL .EQ. NDL) GO TO 170
GO TO 20lC
C CALCULATE PSEDU LOAD
C
425 DO 423 I=I,MM
423 FO(I) ~ Fl(I)-FCJ(I)
IF(ISRE .EQ. 0 .OR. ISRE .EQ. 2) GO TO 442
IF(ISTIFF .EQ. 20) GO TO 940
IF (ISRE .NE. 6) GO TO 951
IF (INAL .LE. 3 .AND. ISTIFF .LE. 4) GO TO 210
GO TO 442
951 IF(ISRE .NE. 3 .AND. ISRE .NE. 4) GO TO 952
IF(ISTIFF .EQ. I .AND. NC .EQ. I) GO TO 210
IF(ISRE .EQ. 3 .AND. NC .EQ. 5) GO TO 210
GO TO 443
952 IF(ISRE .GT. 3) GO TO 953
•
110
99958 FORMAT(3(4X,2Dl3.4,14))
99978 FORMAT (/1 OX, 17HFLEXURAL RIGIDITY, 15X, 14HYOUNGS MODULUS, 20X,
'4HAREA/)
99977 FORMAT(lHl/l0X,8HATCYCLE,14,4X, 2IHLIMITS OF CONVERGENCE ,4X,
• 6HLOWER ,2FIO.5,3X, 6HUPPER,4X,I 4HRESIDUAL FORCE,
• FI1.6) .
99976 FORMAT (/lOX, 34HRE-FORMULATION OF STIFFNESS MATRIX)
99975 FORMAT (3(14,3012.3))
99974 FORMAT (/35X,30H •••••• NODAL STRAINS ETC ••••• , /IX, 4HNODE, 4X,
• 6HSTRAIN, 4X, 12HNEUTRAL AXIS, 2X, 4HCURV, 5X, 4HNODE, 3X,
• 6HSTRAIN, 4X, 12HNEUTRAL AXIS, 3X, 4HCURV, 4X, 4HNODE, 4X,
• 6HSTRAIN, 3X, 12HNEUTRAL AXIS, 3X, 4HCURV//)
99973 FORMAT (/30X, 30H •••• CONVERGENCE AT LOADCYCLE, F8.2, 5H •••• /
• 39X, 20HNUMBER OF ITERATIONS, B/)
99972 FORMAT (/1 OX,22HECCENTRICITY OF THRUST/)
99971 FORMAT (IX,12,F9.5,2F8.0,2FlO.0,F6.3,2Dl2.5,I4)
99970 FORMAT (/IOX,4HNODE, 14,19H HAS DEEMED TOFAIL/IOX,9HAND A HIN,
• 30HGE HAS FORMED AT APPROXIMATELY, F7.2/)
99969 FORMAT (/l5X, 19HTOT AL DISPLACEMENTS/)
99968 FORMAT (/3X,'NODE',2(4X,'Y-AXIS',9X,6HX-AXIS,8X,8HROTA TION,5X) )
99967 FORMAT (3(lX,B,3D 12.5))
99966 FORMAT (lX,I5,3Dl4.6,14)
99965 FORMAT (/23X,5HEND I, 55X, 5HEND 2)
99964 FORMAT (/45X, 15HRESULTANT LOADS/)
99963 FORMAT (/2X,15H VERTICAL FORCE, 6X, IIHHORIZ FORCE, 9X, 6HMOMENT,
• 14X, 14HVERTICAL FORCE, 6X, llHHORIZ FORCE, 9X, 6HMOMENT/)
99962 FORMAT (2(IX,12,2X,Dl2.5,5X,Dl2.5,7X,DI2.5,6X))
8051 FORMAT (2(lX,12,2X,FI2.2,5X,FI2.2,7X, FI2.2, 6X))
99961 FORMAT (//20X, 23H' •• FRAME COLLAPSED"'///)
99960 FORMAT (/20X,"'CONVERGENCE NOT POSSIBLE AT LOAD CYCLE"',16)
99959 FORMAT (/20X,"'EQUATIONS INSOLUABLE AT LOAD CYCLE "',16)
99956 FORMAT (/20X,"'END OF LOADINCREMENTS "', 16)
99957 FORMAT (I5,FIO.5)
99931 FORMAT (/20X,"'INCREMENTAL ••• DISPLACEMENT ••• TOTAL')
8052 FORMAT (60X,'REWIND' REWIND 11',215)
8053 FORMAT (65X,'READ • II ',215)
8054 FORMAT (70X,'WRITE • II',215)
33333 FORMAT(I7X,'STRAIN IN STEEL',/20X,'ES(I,I)',3X,'ESTNI(I,I)',3X,
"ES TN2(1, 1)',3 X,'ES(I,2)' ,3X,'ESTN I (1,2)',3 X,'ES TN2(1,2)')
33334 FORMAT (5X,15,3X,9FIO.6)
END
•
112
COMMON / BLK6/ ESS(I 00, I O),ESP(I 00, I O),EYY(I 00, 1O),EU(I 00, I 0)
*,QLL(300)
COMMON / BLK8 / IQN(I OO),QSL(I 00, I O),QSR(I 00, I O),QD(I 00, I 0)
*,QAS(I 00, I 0)
COMMON/BLKI2/ESS 1,ESP I ,EYYI,EU I
DIMENSION HEAD(5,20),AA(30,3), LODlS(10),NJJD(31),FXD(3I)
*,NJ( I O),NX( 1O),NY(I O),NM( I O),NJJ(I O),FX( I O),FYY(I 0),FMDD(31)
*,FM( I O),NMM( I O),NELX( I O),TX( 1O),AX( I O),BX(I 0),FYYD(31 ),AAD(31 ,3)
*,QL(300),BRX(I 0)
READ(5,11) ((HEAD(I,J),J~1,20),I=I,5)
WRITE (6,12)
WRITE (6, 14)((HEAD(I,J),J~ I ,20),1= I ,5)
READ(5, I) NCONT,NGEOM,M,NDEAD,IDLOAD,ISYMM,ISRE
I ,LODINC,IDLODREV,NCYCREV,NDL,ISSCHK
WRITE( 6,21 )NCONT,NGEOM,M,NDEAD,IDLOAD,ISYMM,ISRE
1 ,LODINC,IDLODREV,NCYCREV,NDL,ISSCHK
NPRINT=O
IPALL ~ 0
IS13 = 5
IS23 = 9
IF(NGEOM.NE.5) GO TO III
READ (5,6) NBE
WRITE(6,21)NBE
READ(5,8) (NMM(I),NELX(I),TX(I),BRX(I), AX(I),BX(I),H,NBE)
WRITE( 6,26)(NMM(I),NELX(I), TX(I),BRX(I),AX(I),BX(I),1= I ,NBE)
READ(5,6) NBE
WRITE(6,21) NBE
READ(5,*) (IQN(I),QL(I),I~I,NBE)
DO 500 IK~I,NBE
IQ=IQN(IK)
READ( 5, *) (QD(IK,I),ESS(IK,I),ESP(IK,I),EYY (IK,I),EU (IK,I),I ~ 1,1Q)
WRITE( 6,56)1 Q,( QD(IK,I),ESS(IK,I),ESP(IK,I),EYY (IK,I),EU (IK,I),
*I=I,IQ)
READ( 5, *)( QSL(IK,I), QSR(IK,I), QAS(IK,I),I ~ 1,IQ)
WRITE( 6,7 6)( QSL(IK,I),QSR(IK,I),QAS(IK,I),I~ I ,IQ)
WRITE(6,57) IK, QL(IK)
500 CONTINUE
GO TO 112
III READ (5,2) T,A,B
WRITE (6,22)T,A,B
GO TO 1110
1110 READ (5,2) BRR
WRITE (6,22)BRR
C DEAD LOAD APPLICATION
112 IF(IDLOAD.NE.3) GO TO 114
READ (5,6) NJLD
WRITE( 6,21 )NJLD
READ (5,7) (NJJD(I),FXD(I), FYYD(I),FMDD(I), I~I,NJLD)
WRITE (6,28)(NJJD(I),FXD(I), FYYD(I), FMDD(I), I~I, NJLD)
114 IF(IDLOAD.NEA) GO TO 113
READ (5,1) LCD,LLD,LTlD,LT2D
WRITE( 6,21 )LCD,LLD,LTI D,LT2D
IF(LCD.GE.l) GO TO 220
113
IF(LLD.GE.I) GO TO 200
220 READ (5,3) ((AAD(I,J), J=I,3),I~I,LCD)
WRITE (6,23) ((AAD(I,J), J=I,3),H,LCD)
DO 221 I~I, LCD
221 AAD(I,3) = -AAD(I,3)
GO TO 113
200 READ(5,3) ((AAD(I,J),J=I,3), I~I,LLD)
WRITE (6,23)((AAD(I,J),J=I,3),I~I,LLD)
DO 201 I~I, LLD
201 AAD(I,2) ~ -AAD(I,2)
113 READ (5,2) DEN, DENI, ROAD
WRITE (6,22) DEN, DENI, ROAD
READ (5,4) FY, YM,FTC, CONVLT, DCONVLT, F1RRC
WRITE (6,24) FY,YM,FTC,CONVLT,DCONVLT, FIRRC
C STEEL PROPERTIES
READ(5,*) ESSI,ESPI,EYYI,EUl
WRITE(6,86)ESS I,ESP 1,EYY 1,EUl
READ(5,*) SAS,SP
WRITE(6,87) SAS,SP
READ (5,6) NJR
WRITE (6,21)NJR
READ (5,6) (NJ(I),NX(I),NY(I),NM(I),I=I,NJR)
WRITE (6,27)(NJ(I),NX(I),NY(I),NM(I),I= I,NJR)
DO 180 I~L3,L4
180 AA(l,2) = - AA(l,2)
130 WRITE (6,31)
GO TO (10,32,20,30),NCONT -~
10 WRITE (6,99999)
GO TO 40
32 WRITE (6,33)
GO TO 40
20 WRITE (6,99998)
GO TO 40
30 WRITE (6,99997)
40 GO TO (50, 60, 70, 80, 81) , NGEOM
50 WRITE (6,99996)
GO TO 90
60 WRITE (6,99995)
GO TO 90
70 WRITE (6,99994) •
GO TO 90
80 WRITE (6,99993)
GO T090
81 WRITE (6,99976)
90 WRITE (6,99992) NDEAD, ISSCHK
99976 FORMAT(/lOX,'FRAME - NGEOM= 5 ')
WRITE (6,99991) LODINC
EEl = 2.0*FYIYM
E ~ BR*P*3/12.0
AR~BR*T
WRITE (6,99989) FY, EEl
WRITE (6,99988) E, AR, A, B, M, YM, DEN, T, BR, DENI,ROAD
IF (NLM .EQ. 0) GO TO 91
IF(LC .GE. 1) WRITE(6,99971) LC,(K,(AA(K,J),J~1,3), K~I,LC)
IF(LL .GE. 1)WRITE(6,99972) LL,(K,(AA(K,J),J=I,2),LODIS(K),K~I,LL)
IF(LTl .EQ. 0) GO TO 91
WRITE (6,99973) L Tl,LT2,(K,(AA(K,J),J~I,3),K~I,L Tl)
WRITE (6,9997 5)(K,(AA(K,J),J= I ,2),LODIS(K),K ~L3 ,L4)
91 WRITE (6,99974)
RETURN
•
•
116 "
DIMENSioN S 11S(3,3),SIZS(3,3),SZlS(3,3),SZZS(3,3)
011=0
O~OABS(O)
ONC~O
ONI~OABS(ONI)
ONZ=OABS(ONZ)
CA=OABS(C)
C
C STIFFNESS AT CENTRE OF ELEMENT
C
C
15 IF (0 .LT. 1.0) GO TO 16
EI ~EC+CA *T/Z.O
EZ~EC-CA *T/Z.O
R~O.O
IF (OABS(E1) .GT. 0.0) R~EZ/E1
E3~E1/EM
E4=(1.0+R)I2.0
E5~(1.0-R)
E6=1.0-E3*E4
CII~E*AA*E6
B 1~-EI*E3*E51T +(QLL(INN)-TI2.0)*CII
CZZ~EI*E6-Z.0*(QLL(INN)- T/Z.)*EI*E3*E51T +(QLL(INN)- T/Z.O)**Z*C II
IF( ISZ3 .EQ. 0) B I ~O.O •
IF(lSZ3 .EQ.I .OR. ISZ3 .EQ. Z) Bl~-BI
IF (C .LT. 0.0) B1~-BI
GO TO ZO
C CRAKEO SECTION
16 El=OABS(C*O*T)
EIM=EI/EM
EZ=Z.0-E1M
E3=QLL(INN)-0*TI2.0
Cll = E*AA*0*EZI2.0
BI = -EI*0*0*E1M/T+E3*Cll
CZZ ~ EI*0**3*EZI2.0-Z.0*E3*EI*0*0*E1M1T+E3*E3*Cl1
IF (ISZ3 .EQ. 0) B I ~O.O
IF (ISZ3 .EQ.I .OR. ISZ3 .EQ.3) B1=-Bl
IF (C .LT. 0.0) BI=-Bl
117 J
C STEEL
20 BIS~O.O
DO 51 JJ~ I ,IQNN(INN)
QXX=AS(lI,JJ)
IF( QXX .LE. 0.0) GO TO 51
QDD= QDX(Il,JJ)
ESS2~ESSN(Il,JJ)
EYY2=EYYN(Il,JJ)
ESP2=ESPN(Il,JJ)
IF(C .LT. 0.0) QDD=T-QDD
ES3~ES(Il,JJ)
IF (DABS(ES3) .GT. EYY2) GOTO 52
CII=CI1+ESS2*QXX
B I S~B I S-ESS2*QXX*(QDD-QLL(INN»
C22~C22+ ESS2 *QXX*( QD D-Q LL(INN»* *2
GO TO 53
52 CII~CI1+ESP2*QXX
B I S=B 1S-ESP2 *QXX*(QDD-QLL(INN))
C22~C22+ESP2 *QXX*( QDD-QLL(INN»* *2
53 CONTINUE
51 CONTINUE
IF (C .LT. 0.0) BIS=-BIS
Bl=B1+BIS
SII ~Cll/XL
SI2~0.0
S13~ -BI/XL
S22= 12.0*C22/XL**3
S23~ 6.0*C22/XL **2
S33= 4.0*C22/XL
72 D=DABS(Dll)
C
C CALCULATE S II IN GLOBAL COORDINATES
C
CC~DCOS(TH)
S ~ DSIN(TH)
C2=CC*CC
S2~S*S
CS=CC*S
SIIS(I,I)=C2*SII+S2*S22 -2.0*CS*SI2
SIIS( I ,2)~CS*(SII-S22) + (C2-S2)*S 12
SI1S(I,3)=CC*SI3-S*S23
S IIS(2,2)=S2*S 11+C2*S22+2.0*CS*S 12
SI1S(2,3)~S*SI3+CC*S23
S 11S(3,3)=S33
S 11S(2, 1)~S 11S(l ,2)
SIIS(3,I)~SIIS(l,3)
S 11S(3,2)=S IIS(2,3)
C
C CALCULATE SI2,S21,S22
C
DO 30 1=1,3
S21 S(l ,I)=-S 11S(l,1)
S21 S(2,1)~-S 11S(2,1)
30 S21S(3,1)~-S IIS(3,1)-SII S(I,I)*XL *S+S 11S(2,1)*XL *CC
DO 40 1~1,3
•
118
DO 40 J=I,3
40 SI2S(l,J)~S2l S(J,I)
DO 50 1=1,3
S22S( I ,I)=-S 12S( I ,I)
S22S(2,I)~-S 12S(2,I)
50 S22S(3,I)~-S l2S(3,I)-S 12S( I ,I)*XL *S+S 12S(2,I)*XL *CC
RETURN
END
SUBROUTINE DECBND(N,BAND,FAIL)
INTEGER FAIL,BAND
C
C ROUTINE TO DECOMPOSE THE TOTAL STIFFNESS MATRIX
SUBROUTINE SLVBND(N,BAND)
INTEGER BAND
C
C ROUTINE TO SOLVE THE DECOMPOSED SET OF EQUATIONS
C
DOUBLE PRECISION SUM,FO,S,D
COMMON/BLK3/ FO (903)
COMMON/BLK4/ S(903,6),D(903)
DO 30 I=I,N
IF (S(I,I) .EQ. 0.0) GO TO 20
SUM ~FO(l)
••••
119
JS ~ I
JE~I-I
IF (JS .LT.I-BAND+I) JS = I - BAND + 1
IF (JE .LT. JS) GO TO 20
DO 10 J=JS,JE
SUM ~ SUM - D(J)*S(J,I-J+I)
10 CONTINUE
20 IF (S(I,I).NE. 0.0) D(I) ~ SUM/S(I,I)
30 CONTINUE
DO 60 II=I,N
I ~ N + 1 -II
SUM = D(I)
D(I) ~ 0.0
C
C IF S(I,I) = 0.0
C
IF (S(I, 1) .EQ. 0.0) GO TO 50
JE~N-I+l
IF (JE .GT. BAND) JE ~ BAND
IF (JE .LT. 2) GO TO 50
DO 40 J~2,JE
SUM ~ SUM - D(J+I-l)*S(I,J)
40 CONTINUE
50 IF (S(I, 1) .NE. 0.0) D(I) ~ SUM/S(I,l)
60 CONTINUE
RETURN
END
~,
120
GO TO 40
30 BI~B-R
Al ~AI2.0
D~CONST-BI
AA ~ GGRAD **2 + 1.0
BB = 2.0*GGRAD*D - 2.0*AI
CC ~ D**2 - BI **2
40 XI ~ (-BB+DSQRT(BB**2-4.0*AA*CC))/(2.0*AA)
X2 ~ (-BB-DSQRT(BB**2-4.0*AA*CC))/(2.0*AA)
YI = GGRAD*XI + CONST
Y2 = GGRAD*X2+ CONST
GGI ~ (DABS(Y I)-YY)/(X I-XX)
G2 ~ (DABS(Y2)- YY)/(X2-XX)
IF (DABS(GGRAD-GG1).LT.0.0000I) GO TO 50
XXX~X2
YYY=Y2
GOT060
50 XXX=XI
YYY~YI
60 YC(KK) = DSQRT«YYY-YY)**2+(XXX-XX)**2)
XL(KK)~ DlST
U(KK) = «DlSTI2.0)**2+YC(KK)**2)/(2.0*YC(KK))
THETA(KK) ~ DATAN(DlST/(2.0*(U(KK)-YC(KK))))*2.0
GA(KK) = DA TAN «Y (L)- Y(l))/(X(L)- X(I)))
GI(KK) ~ GA(KK)+0.5*THETA(KK)
70 CONTINUE
G I (M+ I )~DA TAN«Y(M+ 1)-Y(M))/(X(M+ I )-X(M))) -0.5 *THET A(M)
RETURN
END
DO 80 1~I,M
XL(I)= XYL
X(I+ 1) ~ X(I)+ XX
Y(I+I) ~ Y(I)+YY
GA(I)=GG
80 GI(I) = GG
Gl(M+1) = GG
RETURN
10 CC=-A*A*RI(4.0*B)
DO 20 I = MMQ,M
GAMA ~ PI - THEL* (I-MMQ+I)
GAMA ~DSIN(GAMA) / DCOS(GAMA)
AA ~ 1.0
BB ~ A*A*GAMA/(4.0*B)
X(I) ~ -BB/2.0 + DSQRT(BB*BB/4.0-AA *CC)
Y(I) = B-4.0*B*X(I)*X(I)/(A *A)
20 X(l) = X(I) + A/2.0
GO TO 62
30 CC=A *A *(R-B)*(R-B)-A *A *B*B
DO 40 I ~ MMQ,M
GAMA ~ PI-THEL*(I-MMQ+l)
GAMA = DSIN(GAMA)/DCOS(GAMA)
AA=4.0*B*B+A *A *GAMA *GAMA
BB ~ -2.0*GAMA *A *A *(R-B)
X(I) ~ -BB/2.0 + DSQRT(BB*BB/4.0 - AA *CC)
X(I) ~ X(I)/ AA
Y(I) = B*DSQRT(1.0-4.0*X(I)*X(I)/(A*A»
40 X(I) ~ X(I)+A/2.0
GOT062
50 DO 60 I = MMQ,M
GAMA ~ PI - THEL * (I-MMQ+l)
GAMA = DSIN(GAMA) / DCOS(GAMA)
X(I) = DSQRT(R *RI(l.O+GAMA *GAMA))
Y(I) = DSQRT(R*R-X(I)*X(I» - R+B
60 X(I) = X(I)+A/2.0
62 X(NODE)~A
Y(NODE)=O.O
MMQ=MMQ-2
DO 611=2,MMQ
X(I) = A-X(NODE-I+l)
61 Y(I)= Y(NODE-I+l)
RETURN
END
SUBROUTINE NEWGEOM(J,NGEOM)
C
C ROUTINE TO DETERMINE DEFLECTED GEOMTRIC PROPERTIES
C
GA(J)~ASIN(1.0)
C GA(J) = 3.1416/2.0
IF(DABS(XLL) .GT. 1.00-20) GA(J) = DATAN(YLIXLL)
TC ~ DSQRT (XLL**2+YL**2)
XL(J) = TC
Gl(J) ~ GA(J)
IF(NGEOM .GE. 4)RETURN
U(J) ~ (TC**2/4.0+YC(J)**2)/(2.0*YC(J))
THETA(J) ~ 2.0*DATAN(TC/(2.0*(U(J)- YC(J))))
Gl(J) ~ GA(J) + THETA(J)*0.5
RETURN
END
FO(JJ) ~O.O
70 CONTINUE
IF (LTI .OE. I) 00 TO 140
IF (LC .OE. 1) 00 TO 80
IF (LL .OE. I) 00 TO 100
80 DO 90 K~I,LC
CALL LOAD (AA(K,I), AA(K,2), A, AA(K,3), M, X, MM, BR)
90 CONTINUE
00 TO 190
100 DO 130 N~I,LL
IF(NOEOM .OE. 4 .OR. INO .EQ. I) 00 TO 110
LODI ~ LODIS(N) + I
00 TO (110,120), LODI
110 CALL LOADP (AA(N,2), A, AA(N,I),M,X,MM)
00 TO 130
120 CALL PDIS (XI,YI,AA,NODE,M,N,MM, ROAD)
130 CONTINUE
00 TO 190
140 DO 150 NN=I,LTl
CALL LOAD (AA(NN,I), AA(NN,2),A,AA(NN,3),M,X,MM,BR)
150 CONTINUE
LT4=LTl +LT2
LT3 ~LTl + I
DO IS0NNN=LT3,LT4
IF(NOEOM .OE. 4 .OR. INO .EQ. I) 00 TO 160
LODI ~ LODIS(NNN) + I
00 TO (160, 170), LODI
160 CALL LOADP(AA(NNN,2), A, AA(NNN,I), M, X, MM)
OOTO 180
170 CALL PDIS(Xl,YI,AA, NODE, M, NNN,MM, ROAD)
180 CONTINUE
190 RETURN
END
(
/
.f
125
SUBROUTINE DIS(DI,D2,D3,PI,P,X2,Y2,TH2,XII,X12,YII,
* Y12,C,I,GRAD, ATTH2)
C
C ROUTINE TO DETERMINE LOAD VECTOR ALLOWING FOR DISPERSION
C
SUBROUTINE LOADP(PT,A,P,M,X,MM)
C
C ROUTINE TO EVALUATE LOAD VECTOR FOR POINT LOAD
C
DOUBLE PRECISION A,Fl,FG,PT,AMI,F2,RL,AM2,FGI,P,S,X,FO
COMMON/BLK3/FO(903)
DIMENSION X(301)
S~M
L~O
I~I
10J=I+L
IF (P.LT.x(J)) GO TO 20
L~L+I
GO TO 10
20 FG ~ X(J) - P
RL ~ X(J) - X(J-l)
FGl ~RL - FG
AM2 = PT*FG**2*FGIIRL**2
".'!iiI"
126
.' ..
AMI ~ PPFG I**Z*FG/RL **Z
FZ ~ PT*FG1/RL
FI ~ PT*FG/RL
]]=J-I
11~3*J-Z
JZ ~ 3*J
JJl =3*]] - Z
]]Z = 3*]]
FO(JI) = -FZ + FO(Jl)
FO(12) ~ -AMI + FO(12)
FO(JJl) = -Fl + FO(JJl)
FO(JJ2) ~ AMZ + FO(JJZ)
RETURN
END
SUBROUTINE LOAD(AA,BB,A,W,M,X,MM,BR)
C
C ROUTINE TO EV ALUA TE LOAD VECTOR FOR U,D,L.
C
pt-
O'
127
AMB=-AMA
GO TO 40
100 AR ~ X(J)-BB
XL= X(J)- X(J -I)
Q~-Q
K~O
GOT030
RETURN
END
".:
7HH91 ,PS31 ,QD23,QX23,ESS33,EYY33,ESP33,PM 1,PM2,ES213,ES223,
8HlSl,HIS2,HI61,HI62,HI71,HI72
COMMON IBLK4/S(903,6),D(903)
COMMON 1 BLK61 ESS(IOO,IO),ESP(IOO,IO),EYY(lOO,IO),EU(lOO,l 0)
*,QLL(3 00)
COMMON 1 BLK8 IIQN(IOO),QSL(IOO,IO),QSR(IOO,IO),QD(IOO,IO)
*, QAS(lOO,lO)
COMMON /BLK9/ESTNI (300, 10),ESTN2(300, 10),ES(300, 10)
COMMON/BLK 101AS(300, 10),IQNN(300), QDX(300,1 0)
COMMON 1BLK13/ESSN(300, 10),ESPN(300, 10),EYYN(300, 10),EUN(300, 10)
DIMENSION X(3),W(3),PP(6),WZ(8),Z(8)
C=DCOS(G)
SS=DSIN(G)
11= 3*(19-1)
U1 ~ D(II+I)*C + D(II+2)*SS
U2 = D(II+4)*C + D(II+S)*SS
Vl ~ D(II+2)*C - D(II+l)*SS
V2 = D(II+S)*C - D(II+4)*SS
THI ~ D(II+3)
TH2 = D(II+6)
A6 = A6+(U2-U 1)/XL
A3 ~ A3 + 3.0*(V2_VI)/(XL*XL)-(2.0*TH1+TH2)/XL
A4 = A4+ 2.0*(VI_V2)/(XL*XL*XL)+(THI+TH2)/(XL*XL)
EC ~-A6
D1 ~ 1.0
DNC~ 1.0
CCQ ~ 2.0*A3+3.0*A4*XL
CCQQ=CCQ
CCQNl ~ 2.0*A3
CCQN2 ~ 2.0* A3+6.0* A4*XL
ETTNI = A6-A3*T
ETTN2 = ETTNI-3.0* A4*XL *T
EBBNI ~ A6+A3*T
EBBN2 = EBBN1+3.0*A4*XL*T
ETT = A6-(2.0* A3 + 3.0* A4*XL)*0.S*T
EBB ~ A6+(2.0*A3 + 3.0*A4*XL)*0.S*T
X(l) ~ CCQNI
C STRAIN IN STEEL
DO 300 J=I,IQNN(INN)
ES(I9,J) =0.0
ESTN1(19,J)=0.0
ESTN2(19,J)~0.0
QDEP=QDX(19,J)
ES(19,J)=ETT +CCQ*QDEP
ESTN1 (19,J) = ETTNI+CCQN1*QDEP
ESTN2(I9,J)~ETTN2+CCQN2*QDEP
300 CONTINUE
X(3) = CCQN2
IF(NCONT .GE. 3) GO TO 10
CALL SUBPI (A3, A4, A6, PF1, T, EM, XL, DN1, DNC, DN2
1 ,EMAXN1,EMAXNC,EMAXN2,RR,FM,BR,19,INAL,NC,INN)
."'.,
(j
129
GO TO 40
30 PP(3) = (2.0*FM/EM)*BR*T**3/12.0
PP(6) ~ PP(3)*DABS(CCQN2)
PP(3) ~ PP(3)*DABS(CCQNI)
40 IF(X(I) .GT. 0.0)PP(3) = -PP(3)
IF(X(3) .LT. 0.0)PP(6) = -PP(6)
SUBROUTINE CAL(D,EMX,C,EM,T,A6,QM,FM,I1II,NXX,BR,INAL,NC,INN)
C
C ROUTINE TO EY ALUATE B.MOM. AT ENDS FROM STR, DIST
COMMON 1BLK61 ESS(I 00, I O),ESP( I 00, I O),EYY(I 00, I O),EU(I 00, I 0)
*,QLL(300)
COMMON /BLK9/ESTN I (300, I 0),ESTN2(300, I 0),ES(300, I 0)
COMM ON/BLK I 01AS(3 00, I 0),IQNN(3 00), QDX(300, I 0)
COMMON 1 BLK13/ESSN(300, I 0),ESPN(300, I 0),EYYN(300, I 0),EUN(300, I 0)
EC =-A6
EL= DABS(EC)
130
CA~C
C =DABS(C)
IF(D .LT. 1.0) GO TO 102
C UNCRAKED SECTION
H2=ELIEM
H3=H2/2.0
H4~1.0-H2
H5=1.0-H3
H6=C*T**3/3.0
H7=EL *T*T/2.0
HS~C*C*T**4/(S.0*EM)
H9=( C *T*Q LL(INN))* *2/(4 .O*EM)
Hl O~QLL(INN)/T
HII=1.0-3.0*HI0+3.0*HI0*HI0
HI2=1.0-2.0*HlO
H13~-1.0+4.0*HIO
HI4=3.0-2.0*HlO
QE~2.0*FMIEM
QM=QE*(H4*H6*H ll-HS*H7*H 12-HS*Hl3+H9*HI4)
QM~QM*BR
GOTO 104
C CRAKED SECTION
102 EI~EMXlEM
QM=T*T*FM* (D* D*(2.0*E 1-E 1*E 1)+4.0*D*( QLL(INN)/T -D/2. 0)*
*(3.0*El-EI*El))/12.0
QM=QM*BR
C
C STEEL
104 DO 19 KI~I,IQNN(INN)
QX2~AS(IIII,KI)
QD2~QDX(IIII,KI)
ESS3=ESSN(IIII,KI)
ESP3=ESPN(IIII,KI)
EYY3~EYYN(IIII,KI)
IF (NXX .EQ.I) ES2~ESTNI(IIII,KI)
IF(NXX.EQ.2) ES2~ESTN2(IIII,KI)
IF(CA .LT. 0.0) QD2=T-QD2
IF(DABS(ES2).GT.EYY3) GOTO 29
QM=QM+ESS3 *QX2 *ES2*(QD2-QLL(INN))
GOTOl9
29 H1S=DABS(ES2)
H 16= ESS3 *QX2 *EYY3 *(QD2-QLL(INN))+ ESP3 *QX2 *
*(QD2-QLL(INN))*(HIS-EYY3)
H17~ES2/HIS
QM~QM+HI6*HI7
19 CONTINUE
RETURN
END
C
C ROUTINE TO EVALUATE AXIAL THRUST BY SIMPSON RULE
C
COMMON 1 BLK61 ESS(l 00, I O),ESP(l 00, I O),EYY(l 00, I O),EU(I 00, I 0)
*,QLL(300)
COMMONIBLKIOI AS(300, 10),IQNN(300),QDX(300, I 0)
COMMON 1 BLKI3/ESSN(300, 10),ESPN(300, 10),EYYN(300, I 0),EUN(300, I 0)
DIMENSION P(II)
EC =- A6/EM
EL~DABS(EC)*EM
E=2.0*FM/EM
DO 10 I~I,II
ETl ~ (I-I)*XL/IO.0
C = 2.0* A3 + 6.0* A4*ETl
CA~C
ET = A6-C*TI2.0
EB = A6+C*TI2.0
EMX=99.0
D = 1.0
CALL CALD (C,ET,EB,D,EMX,T,RR,INAL,NC)
IF(I .EQ. 6) RI ~RR
IF(I .EQ. I) DNl=D
IF(I .EQ. I) EMAXNI =EMX
IF(I .EQ. 6) DNC=D
IF(I .EQ. 6) EMAXNC=EMX
IF(I .EQ. II) DN2~D
IF(I .EQ. II) EMAXN2=EMX
IF(D .LT. 1.0) GO TO II
CI~C
C =DABS(C)
HH1=2.0*EM
HH2~T*EL
HH3=C*C
HH4=T*T
HH5~QLL(INN)*QLL(INN)
HH6~C*T*QLL(INN)
P(I)=BR *E*(HH2+HH6-C*HH4/2.0-HH2*ELIHH I-HH3*T*HH5IHH I-HH6*EL/EM-
*HH3*HH4*T/(6.0*EM)+C*HH4*ELIHHI +HH3 *QLL(INN)*HH4IHHI)
C=CI
GOTO 100
II EMX =EMXlEM
P(I) ~ E*BR *T*EM*D*EMX*(3.0-EMX)/6.0
C STEEL
100 DO 90 K=I,IQNN(INN)
QXI2=AS(III,K)
QDEP~QDX(III,K)
ESI=ET+C*QDEP
132
ESS4~ESSN(III,K)
ESP4~ESPN(III,K)
EYY 4~EYYN(III,K)
IF(DABS(ESI) .GT. EYY4) GOTO 91 .
P(I)~P(I)-ESS4*QXI2*ES I
GOT090
91 HH7~DABS(ES I)
HH8=ES S4 *QX 12*EYY 4+ESP4 *QX 12*(HH7 -EYY 4)
HH9=ESIIHH7
P(I)~P(I)-HH8*HH9
90 CONTINUE
10 CONTINUE
PI =P(I)+P(II)+2.0*(P(3)+P(5)+P(7)+P(9))
P 1~PI +4.0*(P(2)+P( 4)+P(6)+P(8)+P( 10))
PI~Pl*XL/30.0
PI=PIIXL
RETURN
END
•
134
APPENDIX -II :,.
FY FIO.3
YM F13.3
FTC FIO.5
CONVLT FIO.5
DCONVLT FIO.5
FIRRC FIO.5
APPENDIX-III
2 5 90 0 2 0 0999900 0 0 0 0
1
1 90 0.5500 0.2540 4.0000 0.0000
1
2,0.275
0.055,200000000.000,0.0,0.00 133,0.20,0.500,200000000.000,0.0, 0.00133, 0.20
0.0,4.000,0.003175000000
0.0,4.000,0.004762500000
23.000 0.000 0.000
27600.00027600000.0003270.0 0.01000 1.00000 0.01000
200000000.000,0.00,0.00133,0.2000
0.0000644,0.190
1
1 1 1 1
1 0
91 0.0000 0.0000 0.1000